GB2559554A - Transmitter and receiver using channel bonding - Google Patents

Abstract

A transmitter includes: a data stream partitioner configured to partition a data stream into two or more stream partitions; two or more framing and interleaving units, each of which is configured to form each stream partition after encoding and interleaving for transmission as frames of data; two or more waveform generators, each of which includes a modulator configured to each receive a stream partition and to generate modulated data from the received stream partition for transmission as two or more of the separate RF channels; and a controller. The controller is configured to generate signaling data and to control the two or more framing units to combine the signaling data with respect to the frames of data for transmission in the one of more separate RF channels. The signaling data is for configuring a receiver to detect and to recover the two or more stream partitions from the two or more RF channels and to recombine the two or more streams partitions into the data stream. The signaling data for each of the RF channels includes an indication of bonding with at least one other RF channel and at least one radio frequency displacement between the radio frequency number of the RF channel and at least one other of the two or more separate RF channels A receiver detects a bonded channel which carries a portioned data stream from another channel which carries another partitioned data stream of a data stream. Accordingly the receiver is still able to detect the data stream even if the RF channels have undergone frequency translation for example if transmitted via translator. The invention has particular application in channel bundling for ATSC 3.0 television broadcast systems.

Description

(71) Applicant(s):

Sony Corporation

1-7-1 Konan,Minato-Ku, 108-0075 Tokyo, Japan (51) INT CL:

H04B 7/02 (2018.01) H04L 25/14 (2006.01) (56) Documents Cited:

WO 2016/091905 A1 US 20140269608 A1 US 20130160069 A1

Rich CHERNOCK, ATSC 3.0: Where We Stand, downloaded 1/8/17 from: https://www.atsc.org/ newsletter/atsc-3-0-where-we-stand/ (72) Inventor(s):

Samuel Asangbeng Atungsiri (58) (74) Agent and/or Address for Service:

D Young & Co LLP

120 Holborn, LONDON, EC1N 2DY, United Kingdom

Field of Search:

INT CL H04B, H04L

Other: WPI, EPODOC and the Internet (54) Title of the Invention: Transmitter and receiver using channel bonding Abstract Title: Channel bonding for transmitting partitioned data streams (57) A transmitter includes: a data stream partitioner configured to partition a data stream into two or more stream partitions; two or more framing and interleaving units, each of which is configured to form each stream partition after encoding and interleaving for transmission as frames of data; two or more waveform generators, each of which includes a modulator configured to each receive a stream partition and to generate modulated data from the received stream partition for transmission as two or more of the separate RF channels; and a controller. The controller is configured to generate signaling data and to control the two or more framing units to combine the signaling data with respect to the frames of data for transmission in the one of more separate RF channels. The signaling data is for configuring a receiver to detect and to recover the two or more stream partitions from the two or more RF channels and to recombine the two or more streams partitions into the data stream. The signaling data for each of the RF channels includes an indication of bonding with at least one other RF channel and at least one radio frequency displacement between the radio frequency number of the RF channel and at least one other of the two or more separate RF channels A receiver detects a bonded channel which carries a portioned data stream from another channel which carries another partitioned data stream of a data stream. Accordingly the receiver is still able to detect the data stream even if the RF channels have undergone frequency translation for example if transmitted via translator. The invention has particular application in channel bundling for ATSC 3.0 television broadcast systems.

At least one drawing originally filed was informal and the print reproduced here is taken from a later filed formal copy.

1/26

11· (Large bitrate) Input stream

Stream partitioner
1 TS/IP input	I TS/IP input
strea^m #1___	stream #m -p—

•10

Tx Upper Layer

ATSC 3.0 modulator #1

ATSC 3.0 modulator #m •12

L

TxPHY

Layer

'.L _ ,1

frequency

TS/IP output stream #m •14

TS/IP output stream #1

	7 f
	Stream re-combiner
k.		J

•14

Terrestrial

Channel

H

Rx PHY Layer

JL .

L

Rx Upper Layer (Large bitrate) Output stream

Fig. 1

2/26

Relative SNR level (db) -5- Relative SNR level (db)

Fig. 2

3/16

ω	CO	σ	co
CL	CL	Φ η	CL
_J		Έί ‘Τ	_l
CL	CL	§ CL	CL

□Q

4/2G

Input GSE packets -Η | 3 I 2 1 11

BBF Creation

SPLITTER

BBF header includes ISCR

	BB	F
	ί 1	2

FEC/

MOD

Input

Stream

Sync.

BB	F
' 3	4

BBF
; 1	2

BBF

FEC/

MOD

ISSY

CNTR

Symbol

Clock

Fig.4

RF 1 RF 2 / \ / \

Fig. 5

6/20

RF1 RF 2

I ^: \

_.60

Fig. 6

7/‘ZG

731

732

734

736

737

738

Large bitrate input stream

>Fig. 7

8/2G

RF 1

RF2

Fig. 8 '82

-841

-842

-843

-844

9/2G

Large bitrate input stream

RF 1 RF 2

I \ / \ —————————-►

Fig. 9

10/2G

100

RF 1 RF m / \ / \

-____________-taFig. 10

11/2G

RF 1 RF 2 ! \ ! \

Fig. 11

12/2

PIP · PLP#ft

FIG. 12

120

04 18

FIG. 13

13/20)

140

142

146

141

143

147

148

Fig. 14

''data

Fig. 15

Fig. 16

14/2G

RF1 RF2

RF3 RF4

a)

173

b)

Fig. 17

15/lG

180

Large bitrate input stream

182

1034

747

748

RF2

RF 1 /

Fig^CT A

16/2.G

RF1 ! \ /

RF2 \

f

191

192

193

194

195

196

190

Fig, 19

17/1G

200

PLP #1 PLP#n

RF 1 RF m

I \ I \

..... >

Fig. 20

18/2G

210

PLP #1 PLP #n

RF 1 RF m / \ / \ ι·ι·ιιι·™··ι·ι I fcrp

Fig. 21

19/2C

RF1

RF2

RF3

RF4

RF5

RF6

Fig. 22

231a232a-

20/2G

PLP #1

PLP#n

Input Processing		Input Processing
1
BICM		BICM
	'f
Time Interleaver		Time Interleaver
	1
Scheduler (Cell mapper)

•231n •232n

-233n

230

235a

RF 1

1	<
Frequency Interleaver	E
		OFDM
	OFDM

235m

RF m

Fig. 23

300

21/ 2.C

Fig. 24

22/2G

310

410

Fig. 25

23/2.6

Fig. 26

24/2&

Fig. 27 lit

From remote control

RF Channel 1

RF Channel 2

Figure 29: DTV receiver for channel bonded programmes

TRANSMITTER AND RECEIVER USING CHANNEL BONDING

TECHNICAL FIELD OF THE DISCLOSURE

The present disclosure relates to a transmitter and a corresponding method for communicating data using at least two separate RF channels. The present disclosure relates further to a receiver and a corresponding method for receiving data using at least two separate RF channels.

BACKGROUND OF THE DISCLOSURE

The “background” description provided herein is for the purpose of generally presenting the context of the disclosure. Work of the presently named inventors, to the extent it is described in this background section, as well as aspects of the description which may not otherwise qualify as prior art at the time of filing, are neither expressly or impliedly admitted as prior art against the present disclosure.

Channel bundling of multiple, separate RF channels to enable total service data rates that exceed the net capacity of a single RF channel is generally known. For instance, channel bundling may be applied above the physical layer, i.e. upper layer solutions take care to split the overall data stream on transmitter side into portions that fit to the capacity of the different single RF channels. Upper layer signaling may be provided that allows the data recombination of the different RF channels in a sorted and stream type consistent way. Channel bundling is processed in a transparent way for all involved RF channels, i.e. the output stream on receiver side is equal to the corresponding input stream on transmitter side. RF channels can be located at any channel frequencies, not necessarily adjacent to each other.

In an ATSC3.0 system it is proposed that each RF channel is handled as a standalone ATSC 3.0 signal. There are no special measures required on the physical layer for bundled channels such as additional guard bands, additional pilots, synchronization etc. Especially the concept of RF channel bundling allows reusing multiple existing LDPC encoder and decoder as well as standard RF tuners, which helps to reduce the overall complexity and simplifies the introduction of high data rate services beyond the capacity of a single RF channel.

In all proposed approaches stream partitioning on transmitter (Tx) side as well as stream recombination (or also denoted as joint BB (baseband packet) de-framing in some figures) on receiver (Rx) side is performed outside the physical layer. However to recover a data stream a receiver must be arranged to identify the RF channels which carry partitions of the data stream.

SUMMARY OF THE DISCLOSURE

Various further aspects and embodiments of the disclosure are provided in the appended claims, including but not limited to, a transmitter, a receiver, a communications device, infrastructure equipment, mobile communications system and a method of communicating.

Embodiments of the present technique can provide an arrangement which allows a receiver to detect a bond channel which carries a portioned data stream from another channel which carries another partitioned data stream of a data stream. For example a transmitter includes a data stream partitioner configured to partition a data stream of data to be communicated into two or more stream partitions, two or more framing and interleaving units, each of which is configured to form each stream partition after encoding and interleaving for transmission as frames of data, two or more waveform generators, each of which includes a modulator configured to each receive a stream partition and to generate modulated data from the received stream partition for transmission as two or more of the separate RF channels, and a controller. The controller is configured to generate signaling data and to control the two or more framing units to combine the signaling data with respect to the frames of data for transmission in the one of more separate RF channels. The signaling data is for configuring a receiver to detect and to recover the two or more stream partitions from the two or more RF channels and to recombine the two or more streams partitions into the data stream. The signaling data for each of the RF channels includes an indication of bonding with at least one other RF channel and at least one radio frequency displacement between the radio frequency number of the RF channel and at least one other of the two or more separate RF channels Accordingly a receiver is still able to detect the data stream even if the RF channels have undergone frequency translation for example if transmitted via translator.

The foregoing paragraphs have been provided by way of general introduction, and are not intended to limit the scope of the following claims. The described embodiments, together with further advantages, will be best understood by reference to the following detailed description taken in conjunction with the accompanying drawings.

BRIEF DESCRIPTION OF DRAWINGS

Embodiments of the present disclosure will now be described by way of example only with reference to the accompanying drawings in which like parts are provided with corresponding reference numerals and in which:

Figure 1 shows a schematic diagram of a current proposal of a channel bundling architecture,

Figure 2 shows a diagram of SNR variations of different terrestrial RF channels,

Figure 3 shows a schematic diagram of channel bundling in DVB-C2,

Figure 4 shows a schematic diagram of channel bundling in DVB-S2x,

Figure 5 shows a schematic diagram of a multi tuner channel bundling receiver architecture,

Figure 6 shows a schematic diagram of a single tuner channel bundling receiver architecture,

Figure 7 shows a schematic diagram of a Tx side processing for channel bundling on cell level with an OFDM symbol,

Figure 8 shows a schematic diagram of the proposed hybrid dual tuner architecture (spreading on OFDM symbol level),

Figure 9 shows a schematic diagram of the Tx side processing for channel bundling on complete OFDM symbol level,

Figure 10 shows a schematic diagram of the Tx side processing for channel bundling on PLP level,

Figure 11 shows a schematic diagram of the proposed hybrid dual tuner architecture (spreading on PLP level),

Figure 12 shows a schematic diagram of the stream partitioner with input and output interfaces,

Figure 13 shows a schematic diagram of the stream combiner for m = 2 with input and output interfaces,

Figure 14 shows a schematic diagram of the Selector/combiner stage for two RF channels,

Figure 15 shows a schematic diagram of an example for exchanging cells of an OFDM symbol in a selector/combiner stage with two RF channels and identical Cdata,

Figure 16 shows a schematic diagram of an example for exchanging cells of an OFDM symbol between a selector/combiner stage with 3 RF channels and identical Cdata,

Figure 17 shows a schematic diagram of two different approaches to interconnect modulators,

Figure 18 shows a schematic diagram of channel bundling with single BICM stage,

Figure 19 shows a schematic diagram of a receiver for channel bundling with single

BICM stage,

Figure 20 shows a schematic diagram of an architecture for channel bundling with multiple PLPs,

Figure 21 shows a schematic diagram of an architecture for channel bundling with multiple PLPs using one broadband frequency interleaver,

Figure 22 shows a schematic diagram of the basic TFS mechanism in terrestrial broadcast systems,

Figure 23 shows a schematic diagram of a Tx architecture for TFS,

Figure 24 shows a schematic diagram of a common transmitter and receiver architecture for use in different modes,

Figure 25 shows a schematic	diagram	of	the	common	transmitter	and	receiver
architecture in ΜΙΜΟ mode,
Figure 26 shows a schematic	diagram	of	the	common	transmitter	and	receiver
architecture in channel bonding mode,
Figure 27 shows a schematic	diagram	of	the	common	transmitter	and	receiver

architecture in MRC mode,

Figure 28 is a schematic block diagram of a transmitter configures to transmit a data stream using channel bonding;

Figure 29 is a schematic block diagram of a receiver for detecting and recovering a data stream from a plurality of bond channel; and

Figure 30 is a schematic block diagram of a three-channel frequency translator.

DESCRIPTION OF EXAMPLE EMBODIMENTS

This disclosure describes an apparatus, method and system for transmitting and receiving data. In some embodiments the disclosure relates to broadcast transmission and reception of data. In some embodiments the data may be audio/video data. Some embodiments describe channel bundling based on multiple RF channels, in particular with SNR averaging across different RF channels.

First, the background and existing architectures will be described, which is related to our copending International patent application WO2016/091005, the contents of which are herein incorporated by reference.

A current proposal for the upcoming ATSC3.0 standard foresees channel bundling of multiple, separate RF channels to enable total service data rates that exceed the net capacity of a single RF channel. Channel bundling is applied above the physical layer in this proposal, i.e. upper layer solutions take care to split the overall data stream on transmitter side into portions that fit to the capacity of the different single RF channels. Upper layer signaling is provided that allows the data recombination of the different RF channels in a sorted and stream type consistent way. Channel bundling is processed in a transparent way for all involved RF channels, i.e. the output stream on receiver side is equal to the corresponding input stream on transmitter side. RF channels can be located at any channel frequencies, not necessarily adjacent to each other.

Each RF channel is handled as a standalone ATSC 3.0 signal. There are no special measures required on the physical layer for bundled channels such as additional guard bands, additional pilots, synchronization etc. Especially the concept of RF channel bundling allows reusing multiple existing LDPC encoder and decoder as well as standard RF tuners, which helps to reduce the overall complexity and simplifies the introduction of high data rate services beyond the capacity of a single RF channel.

In all proposed approaches stream partitioning on transmitter (Tx) side as well as stream recombination (or also denoted as joint BB (baseband packet) de-framing in some figures) on receiver (Rx) side is performed outside the physical layer.

In general the different approaches are explained in the following for two RF channels. Of course it is straight forward to extend the principles to more than two bundled RF channels.

Figure 1 shows the principle of the currently proposed channel bundling mechanism as described in a call for technologies response. PCT patent application PCT/EP2014/061467 and the priority applications on which it is based are hereby incorporated by reference in their entirety.

On the transmitter (Tx) side a stream partitioner 10 partitions the input stream into m TS/IP input streams, which are then separately modulated by m separate modulators 11, 12 into m RF streams for transmission over the terrestrial channel. On the receiver (Rx) side the received RF streams are separately demodulated by m demodulators 13, 14 into m demodulated streams, which are then re-combined by a stream re-combiner 15 into the output stream.

Although the above mentioned solution of the known proposal realizes the simplest solution for channel bundling, it has a significant drawback. As the different channels are completely decoupled, the overall performance of the large input stream suffers from different channel conditions on different RF channels. The main reason is that the transmitter as well as the receiver for every allocated RF channel deploys an own, independent FEC (Forward Error Correction) encoder and decoder. If one or several channels suffer from insufficient SNR or other channel impairments for correct data decoding, the overall recombined stream would still result in a corrupted data stream with errors.

As an example, the plots shown in Figure 2 illustrate measured SNR variations on different terrestrial RF channels from a Swedish operator (Teracom).

Next, existing channel bundling architectures without SNR averaging are described. In particular, existing non terrestrial broadcast systems with channel bundling will be described.

It should be noted first that the channel bundling concept has already been introduced in other broadcasting systems, such as DVB-C2 and DVB-S2x. However in cable and satellite channel no big amplitude changes across the involved RF channels or Data Slices are expected. Therefore an exchange of data cells between the different Tx signal streams for SNR averaging is not required. Nevertheless the principle of the C2 and S2x channel bundling is briefly explained here.

A DVB-C2 system allows spreading data of a single PLP (Physical Layer Pipe) connection over different data slices. This operation mode is intended for advanced services that require throughput rates above the capacity of a single data slice.

All data packets of a bundled PLP connection pass the same input processing block. Inserting the ISSY timestamp in the mode adaptation block allows the reordering of the packets from different data slices on receiver side. At the output of the input processing block the BBFrames of the bundled PLP are spread over the different data slices. Figure 3 illustrates a schematic diagram of a system using channel bundling in DVB-C2.

In DVB-S2x a similar approach (illustrated in Figure 4) is followed as in DVB-C2: A single input stream is carried in parallel over max 3 transponders. As in DVB-C2 the data of a big input PLP passes the same input processing block 40, in ‘High efficiency mode (HEM)’ every BBFrame gets its own ISSY timestamp that allows reordering on Rx side. Splitting is performed in splitter 41 after the BBFrame creation. Figure 4 illustrates channel bundling in DVB-S2x. Every RF channel can use its own PHY parameters, such as symbol rate, modulation and coding settings.

Next, current proposals for terrestrial channel bundling mechanisms as disclosed herein will be described. First, existing channel bundling approaches for terrestrial systems are described.

The transmitter processing is done as described above with respect to existing architectures. For two separate RF channels the related dual tuner receiver architecture is shown in Figure 5, which illustrates a multi tuner channel bundling receiver architecture 50.

It should be noted that decoding of every RF channel is performed independently by separate RF frontends 51, 52 and separate demodulators 53, 53 before joint stream recombination in stream combiner 55 or baseband (BB) de-framing by de-framing unit 56 at the very end of the processing chain. Accordingly a SNR averaging across the two RF channels does not happen.

Some example advantages and disadvantages are stated below. The advantages are:

• Simple and scalable implementation • Complete reuse of separate decoder (tuner + demodulator) • Supports bundling of distant and neighbored channels • Additional statistical multiplex gain across the overall bandwidth The disadvantages are:

• No additional frequency diversity or SNR averaging • No guard band removal possible for neighboring bundled channels (see also below)

For completeness it shall be mentioned that channel bundling can be also realized by an overall bigger, single RF channel. On Tx side the stream is processed by a higher bandwidth input processing, BICM and time interleaver stages.

On Rx side a single broadband tuner can be used, as shown in Figure 6 illustrating a single tuner channel bundling receiver architecture 60 including an RF frontend 61 and a demodulator 62 (including a combined unit 63 for FFT, FDI (frequency domain interleaving), deframing and TDI (time domain interleaving), a QAM-demapper 64, an FDPC-decoder 65 and a BB-deframing unit 66).

Example advantages and disadvantages of this approach are listed below. The advantages are:

• Additional frequency diversity • TFS-like (time frequency slicing) SNR averaging • Additional statistical multiplex gain across the overall bandwidth • Guard band between neighboring bundled channels can be removed The disadvantages are:

• Complex implementation • No bundling of distant channels possible

Next, the proposed channel bundling with SNR averaging will be described. A technology is proposed allowing for SNR averaging across all involved RF channels. Most important element is the spreading of the output data from every FEC encoder (LDPC encoder) across the available RF channels. The spreading may take place in different stages of the TX and RX chain. Two options, namely the spreading on OFDM symbol level and the spreading based on PLP level are described in the following.

First, spreading based on OFDM symbol level will be described. Figure 7 shows an example embodiment 70 of the principle on Tx side for single PLP with channel bundling on two RF channels on cell level within an OFDM symbol. In case of PLP bundling the large input stream is, after input processing in the input processing unit 71, divided in the stream partitioner 72 into different partial streams, each of them allocated to a PLP. These PLPs are then fed into the different modulators 73, 74 and can have the same or different PLP IDs. At least a single partial stream allocated to a single PLP #1 is provided to each modulator. It should be noted in this context that the number of PLPs per RF channel does not necessarily have to be the same. For instance, the first RF channel RF1 might carry n PLPs, while the second RF channel RF2 might carry p PLPs with n^p. However, in case of a single PLP, the dashed boxes shown in Figure 7 are in fact inactive.

In typical architectures such as DVB-T2, DVB-NGH and most likely ATSC3.0, different PLPs are passing different input processing BICM stages as well as individual time interleaver. The BICM (Bit interleaved coded modulation) stage 731, 741 consists of FEC encoder (BCH and LDPC), bit interleaver and QAM mapper. After time interleaving by time interleaver 732, 742 the time interleaved QAM cells of every PLP are then scheduled by scheduler 733, 743 onto different OFDM symbols within a frame.

In a regular case, the output of the scheduler 733, 743 is fed directly to the frequency interleaver and OFDM modulator of the related RF channel. In contrast to current solutions it is proposed to exchange (e.g. equally) the scheduled cells of one OFDM symbol across selected or all involved RF channels. This is performed by selectors 734, 735 and combiner 736 in the modulator 73 and the selectors 744, 745 and the combiner 746 in the modulator 74.

Subsequently, frequency interleaving by frequency interleavers 737, 747 and OFDM modulation by OFDM modulators 738, 748 is performed to obtain the RF output streams.

It should be noted that instead of the shown cell exchange stage and following RF channel frequency interleaver a single frequency interleaver that spans across the overall sum of

OFDM subcarriers or resulting bandwidth of all RF channels (example: single 12 MHz frequency interleaver instead of cell exchange and two separate 6MHz frequency interleavers, see e.g. Figure 21).

Generally, the proposed transmitter comprises the following elements:

i) a data stream partitioner (in the embodiment 70 realized by the stream partitioner 71) 10 configured to partition a data stream of data to be communicated into two or more stream partitions;

ii) two or more modulators (in the embodiment 70 realized by the modulators 73, 74) configured to each receive a stream partition and to generate modulated data from the received stream partition; and iii) an interleaver (in the embodiment 70 realized by the selectors 734, 735, 744, 745, the combiners 736, 746 and the frequency interleavers 737, 747; in other embodiments realized by cell exchange circuitry, i.e. the exchange of cells may also be understood as one embodiment of interleaving in the context of the present disclosure) configured to assign the modulated data generated by a modulator from a received stream partition to different RF channels for transmission.

On Rx side, the carrier exchange from Tx side obviously needs to be reversed before the decoding. Figure 8 shows an example embodiment 80 of the receiver architecture for two RF channels, in particular a proposed hybrid dual tuner architecture (spreading on OFDM symbol level). The disclosure is not limited to two RF channels. The receiver 80 comprises two frontends 81, 82, two demodulators 83, 84, a stream combiner 85 and a de-framing unit 86. Each of said demodulators 83, 84 comprises an FFT and FDI unit 831, 841, a deframing and TDI unit 832, 842, a QAM-demapper 833, 843 and an LDPC-decoder 834, 844. The receiver typically only decodes a single PLP. This is at least the case if one PLP represents one service (e.g. a video stream). In the case that a PLP carries only a service component (e.g. only video or only audio), all PLPs comprising the service need to be decoded.

Generally, the proposed receiver comprises the following elements:

i) a deinterleaver (in the embodiment 80 realized by the FFT and FDI units 811, 821 and 5 the deframing and TDI units 812, 822; in other embodiments realized by cell re-exchange circuitry, i.e. the re-exchange of cells may also be understood as one embodiment of deinterleaving in the context of the present disclosure) configured to receive data of a received data stream via at least two separate RF channels, wherein the data of stream partitions of the data stream are transmitted via the at least two RF channels, and to assign the data belonging to the same stream partition transmitted via different RF channels to different demodulators, ii) two or more demodulators (in the embodiment 80 realized by the QAM-demappers 813, 823 and the LDPC-decoders 814, 824) configured to each receive data of a stream partition and to generate demodulated data from the received data of the stream partition, and iii) a data stream combiner (in the embodiment 80 realized by the stream combiner 82) configured to combine the demodulated data of the two or more demodulators into the data stream.

It is mentioned that the data exchange between the different RF changes can also take place on complete OFDM symbols rather than the above mechanism of cell exchange within an

OFDM symbol. Effectively this means that single OFDM symbols out of the Lp OFDM symbols of a frame (typically data symbols only, excluding preamble and signalling symbols) are exchanged between the different modulators. The related block diagram of such an embodiment 90 of the transmitter is shown in Figure 9, showing Tx side processing for channel bundling on complete OFDM symbol level. It should be noted that the so far cell specific selectors (within an OFDM symbol) 734’, 735’, 744’, 745’ are now working on OFDM symbols within a frame, i.e. the granularity of the selectors 734’, 735’, 744’, 745’ is different compared to the selectors 734, 735, 744, 745 of the embodiment shown in Figure 7, which is denoted by the index in the selectors (Lf instead of Cdata)·

Of course exchange of any form of multiple OFDM symbols is also possible, but comes with less diversity. The exchange of OFDM symbols is seen as important scenario as it allows for a simple combined architecture of channel bundling with TFS with two tuners. It should be noted that the exchange of OFDM symbols may similarly take place after the frequency interleaver.

Next, spreading based on PLP level will be described.

So far it has been assumed that the spreading takes place on OFDM symbol level. This allows for a simple TX and RX implementation but has some disadvantages. The exchanged data rate of the spread OFDM symbols is quite high, since the whole OFDM symbols need to be exchanged, even though the PLP of interest is only transmitted in a subset of the cells of the OFDM symbols. This disadvantage can be overcome by spreading the cells on PLP level, as depicted in Figure 10, showing an embodiment 100 of the Tx side processing for channel bundling on PLP level.

The selection of data cells in the different selector blocks 1021, 1022, 1031, 1032 is for example defined by the partitioning rate in the stream partitioner 101, i.e. the relation of the BB Frames for this PLP that are fed into the different modulator chains 102, 103. In this embodiment the schedulers 1024, 1034 are provided after the combiners 1023, 1033.

In this embodiment it is also shown that separate input streams can be handled by separate input processing units and separate stream partitioners. Alternatively, separate input streams may be handled by a common input processing unit 71 and a common stream partitioner 101. The proposed idea can thus be applied to handling separate input streams in parallel as well.

The respective receiver architecture 110 is shown in Figure 11 showing the proposed hybrid dual tuner architecture (spreading on PLP level) comprising two demodulators 113, 114. The demodulators 113, 114 only exchange the LLR values of the QAM demappers 833, 843, resulting in a smaller data rate compared to the exchange of OFDM symbols as provided in the receiver architecture 80 shown in Figure 8. Note that different possibilities exist for the exchange of data cells on reception side: On top to the illustrated exchange of LLR values after the QAM demappers 833, 843 per bit, it is also possible to exchange I and Q values as well as the channel state information (CSI) by the common units 1131, 1141 for FFT, FDI, demapping and TDI before the QAM demappers 833, 843.

Another advantage of exchanging the data on PLP level is the increased flexibility in case of 5 channel bundling with different RF bandwidths, which leads to different OFDM symbol durations. While there are difficulties exchanging OFDM symbols, due to the different

OFDM symbol timing, this causes no problems in case of exchanging cells on PLP level. It should however be ensured that the number of exchanged cells between the different RF channels is adapted according to the capacity of the different channels.

Next, a comparison of channel bundling with SNR averaging with state of the art channel bundling approaches will be made. Example advantages are:

• Relatively simple and scalable implementation • Reuse of existing tuners and almost completely existing demodulators • Supports bundling of distant and neighbored channels · Additional frequency diversity • TFS-like (time frequency slicing) SNR averaging • Works for single PLP as well as for multiple PLPs across multiple RF channels Example disadvantages are:

• Requires high data rate interface between demodulator chips (received QAM cells + channel state information or LLR values after QAM demapping for PLP(s) of interest) • No guard band removal possible for neighboring bundled channels

The skilled person will appreciate that in some systems received signals from different reception antennas and tuners are combined into a single decoder chip.

Next, a detailed description of example embodiments of the stream partitioner, the stream combiner and the selector will be provided.

The input streams of the stream partitioner 120, as depicted in Figure 12 showing an embodiment of a stream partitioner 120 with input and output interfaces, consist of baseband frames (BB-Frames) of the n different PLPs. Each PLP may have a different input stream format, such as TS, IP or GSE. The corresponding input stream packets are packetized to BBFrames with a suitable timestamp such as an ISSY timestamp in the input processing blocks 121, 122 before the stream partitioner 120. The task of the stream partitioner 120 is to distribute the BB-Frames of the n PLPs to m streams in such a way, that the output data rate of the streams designated for the m modulators matches the available capacity of the corresponding RF channel. In the simplest case with m = 2 modulators with the same capacity, the stream partitioner equally partitions the input streams to both modulators. In case of different transmission parameters or RF channel bandwidths of the m RF channels, the capacity of the m RF channels may however be different, requiring an uneven output stream capacity distribution at the stream partitioner output.

The task of the stream combiner 130 in the receiver, depicted in Figure 13 showing an embodiment of a stream combiner 130 for m = 2 with input and output interfaces, is to revert the process of the stream partitioner. For a given PLP that is being decoded the m streams of the m demodulators 131, 132 are joined according to the ISSY timestamps available in the BB-Frames. The stream combiner 130 comprises a buffer to store the BB-Frames from the different streams, whose size depends on the maximum difference decoding delay of the m demodulators. In case of identical decoding delay a buffer size of some BB-Frames is sufficient. The processing of the partitioner and combiner is transparent, i.e. the stream of BBFrames at the output of the stream combiner 130 is identical to the input of the stream partitioner. After the stream combining the initial TS/IP/GSE stream is restored by the BB deframer 133.

The task of the selector and combiner stage 140, depicted in Figure 14 showing an embodiment of a selector/combiner stage 140 for two RF channels, is to equally distribute the modulated symbols (so called cells in DVB) at the output of the m schedulers 141, 142 across all m RF channels. This way the frequency diversity is increased from the bandwidth of a single channel by a factor of m to the overall bandwidth of all bundled channels. In particular in case of different SNR levels of the RF channels, an SNR averaging takes place across all

RF channels. It should be noted that selector(s) 143, 144, 145, 146 provided in each modulator may be implemented as separate selector blocks as e.g. shown in Figure 9, i.e. one selector block per modulator so that for m modulators each modulator comprises m selector blocks. In another embodiment each modulator comprises a single selector unit performing the functions of the selector blocks. The data received from the selectors 143, 144, 145, 146 are combined by combiners 147, 148.

The simplest way of applying such a distribution is the exchange of each m-th cell (of all Cdata cells) of each OFDM symbol between all involved RF channels. For the simplest case of m = 2 with the same RF bandwidth (i.e. Cdata is identical for all RF channels), all cells of the

OFDM symbols with even index (index 2:2: Cdata, according to MATLAB syntax) remain in the current RF channel, while all cells of the OFDM symbols with odd index (index 1:2: Cdata, according to MATLAB syntax) are exchanged between both RF channels. The block diagram of such a selector/combiner stage 140 is shown in Figure 14. The impact to an OFDM symbol is depicted in Figure 15. Figure 15 shows an example for exchanging cells of an OFDM symbol in a selector/combiner stage with two RF channels and identical Cdata· A similar example with three RF channels is shown in Figure 16 showing an example for exchanging cells of an OFDM symbol between a selector/combiner stage with 3 RF channels and identical Cdata·

In the more general case of m RF channels with different bandwidths and hence different

OFDM symbol durations and number of cells Cdata, the amount of exchanged cells is not equal and must be calculated according to the ratio of Cdata of the different RF channels as well as the different OFDM symbol durations. Further, the order of the combining operation of the cells from the different RF channels must be defined. Both numbers have to be defined in a deterministic way using rounding operations to avoid unambiguities between the modulator and demodulator implementation. To ensure a random distribution of the cells per RF channel after the combining stage, a frequency interleaver is applied individually for each RF channel.

Next, the number of required communication links between modulators/demodulators will be discussed.

In case of m = 2, four unidirectional communication links, or two bidirectional communication links are necessary to exchange the data between the m selector/combiner stages. With increasing number of m, the number of required communication links grows quite fast. This holds for both the transmitter (comprising m modulators) and receiver (comprising m demodulators). To avoid a large number of dedicated transmission links or better logical representation, a communication bus between the m demos may be used. The two approaches of using dedicated links or a communication bus are exemplarily shown in Figure 17. Figure 17 shows two different approaches to interconnect several modulators 171, 172, 173, 174 (Figure 17(a) shows dedicated links, Figure 17(b) shows a communication bus).

The BB-Frame input and output streams of the selector/combiner stages are omitted for simplicity.

Next, channel bundling with single BICM stage will be described.

It will be described briefly how channel bundling is realized with a single BICM encoding and decoding stage (i.e. FEC encoding (BCH/LDPC), bit interleaving and QAM mapping). In contrast to the previous solutions that focus on reusing existing functional blocks or even existing whole demodulator architectures, this option requires that the BICM stages can handle data rates beyond the capacity of a single RF channel, however the architecture for the transmitter and receiver gets more simple.

An embodiment of the transmitter side architecture 180 with single BICM stage and multiple bundled RF channels is shown in Figure 18 for an example configuration of two RF channels. An embodiment of the related receiver block diagram 190 is shown in Figure 19 for channel bundling with single BICM stage.

The transmitter 180 comprises two modulators 181, 182, wherein the modulator 181 comprises a splitter 183 rather than selectors and a combiner as the provided in the embodiment of the transmitter 100 shown in Figure 10. Thus, rather than using the same complete BICM chain another option is provided: If the splitter is located after the same FEC (LDPC) encoder, every partial bitstream into the different modulator chains can be modulated by separate QAM modulators and allow therefore for different robustness levels on the different RF channels. On receiver side the combination then takes place after passing separate QAM demappers. For this purpose the receiver 190 comprises separate frontends 81, 82, separate units 191, 192 for FFT, FDI and deframing and a single combiner 193, a single time deinterleaver 194, a single QAM demapper 195, a single LDPC decoder 196 and a single deframing unit 86.

Next, the generic architecture for multiple PLP and the relation to time frequency slicing will be described.

A proposed enhancement focuses on channel bundling for very high data rate single PLPs. However, in general the exchange of the subcarriers from different encoding chains is also applicable to a multiple PLP scenario. Of course mixed scenario with a high data rate PLP allocating almost the overall capacity and other PLPs to fill the remaining capacity are possible. The Tx structure for M-PLP with n PLPs and m RF channels is shown in Figure 20 depicting a transmitter architecture 200 for channel bundling with multiple (i.e. n) PLPs. Compared to the transmitter architecture 70 shown in Figure 7 n input processing units 71a,

..., 71n are provided for separate input processing of the n PLPs. Further, each of the m modulators 73a, ...,73m comprises n BICM units 731a, ..., 731n and 741a, ..., 741n, n time interleavers 732a, ..., 732n and 742a, ..., 742n, a scheduler 733, 743, selectors 734, 735, 744, 745, a combiner 736, 746, a frequency interleaver 737, 747 and an OFDM modulator 738, 748.

Instead of the cell exchange stage and following RF channel frequency interleaver, a single frequency interleaver spanning across the resulting bandwidth of all RF channels can be used, obviously still requiring an interface between the m modulators. This is shown in Figure 21 depicting a transmitter architecture 210 for channel bundling with multiple PLPs using one broadband frequency interleaver 212 for all modulators 211a, ..., 21 In and separate input processing units 213a, ..., 213n, 214a, ..., 214n for each PLP in said modulators 211a, ..., 211n.

One drawback of the proposed channel bundling approach is the mandatory usage of several tuners. It should be mentioned that there is another proposal in DVB and ATSC3.0 that allows for a single tuner reception of several RF channels, the so called “Time Frequency Slicing (TFS)”. TFS also spreads data from different PLPs to different RF channels (up to 6 frequencies) to make a single 'virtual' channel to allow efficient statistical multiplexing. PLPs are scheduled that they appear only at one RF channel at one point of time. Some guard band between different PLP portions on different RF channels is provided to enable channel change. Figure 22 shows the basic TFS mechanism in terrestrial broadcast systems (DVB, ATSC3.0).

However, there are also drawbacks and constraints in using TFS, such as the mandatory usage of a sufficient number of PLPs to allow for enough time for RF channel hopping. Moreover it is not possible that a PLP allocates the majority of the available capacity as this would not allow to have this PLP scheduled to a single RF channel at all possible times. This results in constraints regarding the maximum capacity of one PLP, which is significantly below the capacity of a single RF channel. Figure 23 shows a Tx architecture 230 for TFS which may be compared to the Tx architecture 180 for channel bundling shown in Figure 18. This transmitter 230 comprises n processing chains for processing the n PLPs, each processing chain comprises an input processing unit 231a, ..., 23In, a BICM 232a, ..., 232n and a time interleaver 233a, ..., 233n. Further, a common scheduler 234 is provided. The output of the scheduler 234 is provided to m separate OFDM units 235a, ..., 235m, each including a frequency interleaver and an OFDM modulator.

Generally speaking TFS cannot transmit at data rates above the capacity of a single channel. This is extended by channel bundling, using for example the methods and apparatus described.

In other embodiments of the present disclosure, the proposed transmitter and receiver architecture is used and may be embodied as a unified architecture, in other scenarios in which two tuners are used in the receiver architecture. Such other scenarios include - in addition to the above explained scenario using channel bundling (also called channel bonding) - ΜΙΜΟ architectures and MRC (Maximum Ratio Combining) architectures used in diversity receivers. This does not preclude use adaptation for other scenarios. In some embodiments, standard blocks from a standard SISO receiver, including BICM stages, shall be reused. The proposed architectures use a joint processing / cell exchange stage across two (or more) transmitter modules and across two (or more) receiver modules, respectively. The proposed architectures provide the advantages of a lower implementation and development effort, lower costs because of reused blocks and therefore higher success chances in the market.

It should be noted that LDPC decoders as most complex elements of the receiver are implemented in parallel structures: Logically two separate LDPC decoders with processing speed x could be also handled by a single LDPC decoder operated at processing speed 2x. In other words: Implementing with two standard speed LDPC decoders or double speed single decoder is functionally equivalent.

A common architecture of a transmitter 300 and of a receiver 400 is shown in Figure 24. Elements that are already explained above with reference to other embodiments will be provided with the same reference numerals as in those other embodiments. In the transmitter 300, a unified precoding and cell exchange unit 301 is provided; in the receiver 400 a unified decoding and cell re-exchange unit 401 is provided. Further, in the receiver 400 OFDM demodulators 402, 404 and common units 403, 405 for FDI, PLP selection and TDI are provided.

As can be derived from the schematic diagram, the processing will be made on symbol level (e.g. on the level of QAM symbols), which is common to all applications of this common architecture. Further, an interface for requesting and/or obtaining redundancy data (e.g.

redundancy data on demand via a separate channel) may be provided based on the same processing level.

Since the transmitter can be operated in different modes (i.e. MRC mode, ΜΙΜΟ mode and channel bonding mode) a control unit 302 is provided in some embodiments to control the unified precoding and cell exchange unit 301 accordingly to work in the desired mode. This control unit 302 may be operated by the operator of the transmitter. The control unit may select ΜΙΜΟ operation for a certain RF channel if the transmitter is equipped with several antennas or may operate in channel bonding mode for other two RF frequencies. The operation may depend on the network design and anticipated receiver capabilities and are selected by the network operator. Further, some signaling is included into the transmitted data streams identifying the respective mode in which the transmitter 300 is operated for use by the receiver 400 so that the receiver 400 can operate the unified decoding and cell re-exchange unit 401 in the same mode. The signaling may be embedded for example in layer 1 signaling which may be carried in a preamble or signaling symbols at the beginning of each frame defining the mode of operation and is used by the receiver to decode the following data part depending of the mode of operation.

Figure 25 shows the transmitter 310, corresponding to transmitter 300, and the receiver 410, corresponding to receiver 400, when operated in the ΜΙΜΟ mode. In this case the unified precoding and cell exchange unit 301 functions as ΜΙΜΟ encoder 311 and the unified decoding and cell re-exchange unit 401 functions as ΜΙΜΟ decoder 411.

In the ΜΙΜΟ encoder 311 a linear precoding matrix may be used. Further, a different precoding matrix may be applied per subcarrier k. The precoding may use eSM (enhanced Spatial Multiplexing) and PH (Phase Hopping) which may have the following precoding matrix per subcarrier k ¥(k)

1 0 i	' cost/?	sint/f
Q ρ 5#lA) j		CQS1p_

Additional precoding elements, like power allocation or stream-based phase hopping may be used additionally. Precoding at the transmitter increases diversity and improves the overall system performance. In another embodiment plain spatial multiplexing may be applied with

1.J

In this case, no precoding is applied and the precoder can be considered to be transparent.

The RF channels RF1 and RF2 are defined in the spatial domain, and the two tuners are connected to two antennas For performing ΜΙΜΟ transmission at least two transmit antennas and at least two receive antennas are provided, i.e. a first transmit antenna transmits data on

RF1 and a second transmit antenna transmits data on RF2. There may be interference between the transmit antennas and the receive antennas. The channel matrix may be represented as ο ^Λ11 ^12

Λζΐ ^22Ιη the receiver 410 ZF (Zero Forcing) or MMSE (Minimum Mean Square Error) detection may be used for decoupling the two received data streams. Alternatively, a ML (Maximum Fikelihood) demapper may be used, for example a joint ΜΙΜΟ decoder and QAM demapper

412.

Figure 26 shows the transmitter 320, corresponding to transmitter 300, and the receiver 420, corresponding to receiver 400, when operated in the channel bonding mode. In this case the unified precoding and cell exchange unit 301 functions as cell exchange unit 321, representing another embodiment of the disclosed interleaver, and the unified decoding and cell re-exchange unit 401 functions as cell re-exchange unit 421, representing another embodiment of the disclosed deinterleaver,. This case may also be seen functionally as a subset of the ΜΙΜΟ case.

In the cell exchange unit 321 SNR averaging may be described in matrix notation to highlight the analogy to the ΜΙΜΟ description

Y(even) - [θ θ] and V(odd) - [θ θ].

Hereby, even and odd refer to OFDM carrier numbers, but could also map to other granularities (OFDM symbols, ...). In another embodiment any precoding (similar to ΜΙΜΟ) may be applied providing increased diversity. However, unitary precoding, similar to the ΜΙΜΟ precoding described above, is preferred, but not essential to the disclosure. Compared to the simple cell exchange described by the matrices above, precoding further improves the performance. If the precoding block is already available for the ΜΙΜΟ operation mode, the precoding block may be used for the channel bonding mode without additional complexity. The improved performance by applying precoding may be explained as follows: precoding overlaps the two symbols generated by the two BICM chains, i.e. a superposition of both symbols is transmitted in each RF channel. If the two RF channels experience very different fading or attenuation, the information of the two symbols may be recovered more reliably at the receiver (in an extreme scenario, the information of both symbols may be recovered from only RF channel; if no precoding is applied, the second symbol may be lost if one the two RF channels is attenuated very strongly).

The RF channels RF1 and RF2 are described in the frequency domain. There is no co-channel interference. The channel matrix may be represented as where hll and h22 correspond to the fading coefficients of the RF channels RF1 and RF2, respectively. This description is intended to highlight the analogy to the ΜΙΜΟ mode above.

In the receiver 420 cell re-exchange is performed, e.g. by use of reordering of SNR averaging or inverse precoding.

Figure 27 shows the transmitter 330, corresponding to transmitter 300, and the receiver 430, corresponding to receiver 400, when operated in the MRC combining mode providing a diversity receiver. In this case the transmitter 330 uses only a single path (modulator), i.e. there is only a single data stream processed. The unified precoding and cell exchange unit 301 is disabled. In the receiver the unified decoding and cell re-exchange unit 401 functions as cell re-exchange unit 431 (representing still another embodiment of the disclosed deinterleaver), but issues a single data stream as well so that only one path after the cell reexchange unit 431 is activated. This case may also be seen as a subset of the ΜΙΜΟ case.

The RF channel between the transmission path and each receiving path may be represented as /ill ^'21

In the receiver 430 maximum ratio combining may be performed, e.g. by use of coherent addition of the two received signals following the algorithm of maximum ratio combining (MRC). More advanced combing algorithms may be used as well, e.g. optimum combining.

Channel Bonding Embodiments

As indicated above, embodiments of the present technique can provide an arrangement for bonding channels in a way which is not affected by a frequency translation. As explained above a plurality of channels, which carry one or more PLPs may be bundled together to increase a bandwidth for a service with respect to a maximum bandwidth that can be provided with one channel. This is because the capacity of a single RF broadcast channel of 6, 7 or 8MHz is rather limited to around 50Mb/s. Therefore for broadcasting of high bit rate signals such as those needed for UHDTV, channel bonding can be used. In channel bonding, a high bit rate programme stream is partitioned into two or more lower capacity partial streams. Each partial stream is then delivered to its own exciter or modulator each of which broadcasts its particular partial stream via a separate RF channel. Figure 28 provides a simplified representation with respect to the examples presented above with reference to Figures presented above in which like parts have the same numerical references. The schematic block diagram of Figure 28 has been reproduced substantially from the ATSC3.0 specification [1], As shown in Figure 28, the stream partitioning 72 provides on two output channels 2801, 2802 two bit streams to be carried via two bonded channels. The two bit streams are fed to BICM units 731, 741, before then being fed to separate framing and interleaving units 2804, 2806. The outputs from the framing and interleaving units 2804, 2806 are fed to waveform generation units 2810, 2812 which are used to generate two RF channels as explained in the above mentioned embodiments. Optionally the transmitter includes a cell exchange unit 2814 to perform interleaving of cells between the two channels as explained above.

To view each such channel-bonded programmes, the receiver requires as many demodulators as there were partial streams at the transmitter. The partial streams output from all the demodulators have to be re-combined into a single programme stream for input into the media player of the receiver. The ATSC3.0 specification A/322 [1] supports channel bonding across two RF channels, the contents of which are herein incorporated by reference.

According to example embodiments, the number of channels which are bonded is two. As explained above with reference to the example embodiments presented in Figures 1 to 27, a receiver for receiving two-channel bonded programmes needs two tuners and two demodulators as illustrated in Figure 29 in which like parts have the same numerical references. As shown in Figure 29 an RF antenna 2901 detects radio signals which are fed to two receiver chains each of which includes a tuner 2902, 2904 (A and B) and a demodulator 2906, 2908. A controller controls the tuners 2902, 2904 to detect and to tune each of the tuners 2902, 2904 to isolate respective RF channels carrying the partial streams of bits of the high bandwidth service. A stream combiner 2912 then re-combines the respective bit streams into an estimate of the original stream which is then output via a multiplexer 2914 to a media processor 2916 generating text/data and audio/video content 2918, 2920.

As shown in Figure 29, the controller 2910 receives from demodulator A 2906 Layer 1 signalling data recovered from a first channel, which provides signalling data which includes a channel bonding flag for controlling the multiplexer 2914.

When one of its principal demodulator 2906 starts to demodulate a given RF channel, the receiver controller 2910 needs to know whether or not the stream carried on the particular RF channel is a full or partial stream. If the stream is a full programme stream, the demodulator 2906 output is delivered directly to the media player 2916. If the stream is a partial stream, the demodulator 2906 needs to, in addition find out which RF channel carries the second partial stream. This information is carried in the Layer 1 (LI) signalling of each bonded RF channel. A demodulator finds this information by decoding the LI signalling and has to pass this information to the controller 2910 of the receiver. The controller 2910 in-turn instructs the second tuner 2904 and demodulator 2908 to tune to the second RF channel. The partial streams output from the two demodulators 2906, 2908 go into the stream combiner 2912 which combines them to form a syntactically correct programme stream which is then delivered to the media processor 2916 as illustrated in Figure 29.

Broadcast networks often use channel transposers or translators to extend the coverage area of the main transmitter station. A channel translator does not demodulate the DTV signal - it simply translates an RF channel of a particular absolute radio frequency number (ARFN) to an RF channel of another ARFN which it then retransmits. Figure 30 is a schematic of a threechannel frequency translator, which translates RF channels at input frequencies Frto output RF channels at output frequencies F°k. In this example, k can be 0, 1 or 2. As shown in Figure 30, an RF antenna 3001 detects radio signals which are fed to three respective tuners 3002, 3004, 3006. The respective outputs from the tuners 3002, 3004, 3006 are fed to amplifiers

3010, 3012, 3014 which output amplified signals to amplitude modulators 3016, 3018, 3020. According to this arrangement, for the k-th input RF channel, the tuners 3002, 3004, 3006 tune to a frequency Fr and down-converts the detected RF signal to an intermediate frequency where it is filtered , amplified by the amplifiers 3010, 3012, 3014 and then amplitude modulated by the amplitude modulators 3016, 3018, 3020 to output frequency F°k. If the translator is a multi-channel translator as in Figure 30, the re-modulated signals once filtered by filters 3030, 3032, 3034 are then combined by a combiner before passing the multi-channel signal through a high power amplifier HP A 3038 and then transmitted by a transmit antenna 3040.

Frequency translators present a particular problem to bonded channels. Let the two bonded channels enter the frequency translator on ARFNs Βχ and B₂. The LI signalling of the RF channel with ARFN = Bi carries B₂ as the ARFN of its bonded channel. Equivalently, the LI signalling of the RF channel with ARFN = B₂ carries Bi as the ARFN of its bonded channel. Consider that the translator translates RF channel Bfio B₄ and B₂ to B₆. When the two RF channels (B₄ and B₆) reach the receiver, it would not be able to find the bonded channel. Let the first demodulator tune to B₄. From its decoded LI signalling the controller will detect that the second bonded RF channel ARFN = B₂ but since the translator had moved this channel to B₆, the second tuner will not find the required partial stream at B₂.

Embodiments of the present technique can provide an arrangement in which bonded channels can be signalled to a receiver in a manner that is not affected by frequency translation. Embodiments of the present technique can therefore provide an arrangement in which:

1. Instead of carrying the ARFN of the second RF channel in the LI signalling (as described in section 9.3.6 of [1]), carry instead the absolute radio frequency displacement (ARFD) between this channel and its bonded RF channel. In the case illustrated above, the LI signalling of the RF channel with ARFN = Bi will instead carry (B₂-B₄) as the ARFD of its bonded channel. Equivalently, the LI signalling of the RF channel with ARFN = B₂ carries (B₄-B₂) as the ARFD of its bonded channel. Then at the receiver, the controller knowing the ARFN of the first tuner-demodulator, can use the ARFD decoded from its LI signalling to compute the ARFN of the second bonded channel. Example: knowing that the ARFN is B₄ and reading (B₂-B₄) from its

LI signalling, the ARFN of the second bonded channel can be computed as (Bi+(B₂Bi) = B₂).

2. The translators have the ability to enforce a configured ARFD between a configurable pair of input RF channels. This means that for each input ARFN needing translation, the design should allow configurability based on two parameters: {Paired ARFN, ARFD} where:

• Paired ARFN is the ARFN of the RF channel with which the input ARFN is bonded • ARFD is the absolute radio frequency displacement that must be maintained between this channel and its bonded RF channel by the translator.

For any given input RF channel, a configuration in which {own ARFN, 0} (i.e. Paired ARFN is set to input ARFN and riW/J = 0) designates that the RF channel may be translated independently i.e. without enforcement of a frequency displacement to any other RF channel. On the other hand, for any two input RF channels B_n and B_m with respective configurations {B_m, D_n=(B_m-B_n)} and {B_n, D_m=(B_n-B_m)}, the translator can discern the following:

• The two channels are bonded - given that Paired ARFN = B_m in the configuration of B_n and that Paired ARFN = B_n in the configuration of B_m • The frequency displacement to enforce at the outputs is |B_n-B_m| • B_n should be translated to a higher ARFN than B_m since D_m is positive and D_n is negative.

Therefore, the defined configuration information is enough for the translator to maintain the necessary frequency displacement between the two bonded channels at its output. Since the translator can so ensure that the ARFD between input pairs of bonded channels is also respected at the translator output, it does not matter to the receiver that two bonded and translated RF channels are no longer on their original ARFN. Taking the translation example above, the translator ensures that (B₆-B₄) = (Β₂-Βχ) and so a receiver having tuned to B₄ on its primary tuner-demodulator, and reading (B?-B i) from the LI signalling, the controller can compute the ARFN of the second bonded RF channel as:

(Β₄+(Β₂-Βχ)) B₄+(B6-B₄) = Βθ.

The LI signalling parameter ARFD has the same dynamic range as the ARFN and so requires the same number of bits.

The configuration for each translatable input RF channel is expected to be set up when the particular translator is installed and changed as necessary whenever the frequency plan is changed.

In another embodiment, bonding can be done intra-channel: one partial stream is transmitted on the vertical polarisation and the other on the horizontal polarisation of the same RF channel. In this case the LI signalling parameter ARFD is set to zero (there is zero frequency displacement between the bonded ‘channels’) and the LI signalling parameter that signals that channel bonding is active is set to indicate channel bonding. Accordingly, a receiver controller whose primary tuner-demodulator decodes ARFD = 0 and channel bonding active from its LI signalling would instruct its second tuner-demodulator to tune to the same RF channel but with a polarisation opposite to that used by the primary tuner-demodulator. Equivalently, the configuration for such ‘bonded polarisation’ RF channels in the translator can be set to {own AFRN, 0}. This informs the translator to treat this RF channel as if it was not bonded.

Thus, the foregoing discussion discloses and describes merely exemplary embodiments of the present disclosure. As will be understood by those skilled in the art, the present disclosure may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. Accordingly, the disclosure of the present disclosure is intended to be illustrative, but not limiting of the scope of the disclosure, as well as other claims. The disclosure, including any readily discernible variants of the teachings herein, defines, in part, the scope of the foregoing claim terminology such that no inventive subject matter is dedicated to the public.

In the claims, the word comprising does not exclude other elements or steps, and the indefinite article a or an does not exclude a plurality. A single element or other unit may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

In so far as embodiments of the disclosure have been described as being implemented, at least in part, by software-controlled data processing apparatus, it will be appreciated that a nontransitory machine-readable medium carrying such software, such as an optical disk, a magnetic disk, semiconductor memory or the like, is also considered to represent an embodiment of the present disclosure. Further, such a software may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems.

The elements of the disclosed devices, apparatus and systems may be implemented by corresponding hardware and/or software elements, for instance appropriated circuits. A circuit is a structural assemblage of electronic components including conventional circuit elements, integrated circuits including application specific integrated circuits, standard integrated circuits, application specific standard products, and field programmable gate arrays. Further a circuit includes central processing units, graphics processing units, and microprocessors which are programmed or configured according to software code. A circuit does not include pure software, although a circuit includes the above-described hardware executing software.

The following numbered paragraphs provide further example aspects and features of the present technique:

References [1] ATSC Standard: Physical Layer Protocol (A/322)

Claims (4)

1. A transmitter for transmitting data using at least two separate RF channels, the transmitter comprising:

a data stream partitioner configured to partition a data stream of data to be communicated into two or more stream partitions, two or more framing and interleaving units, each of which is configured to form each stream partition after encoding and interleaving for transmission as frames of data, two or more waveform generators, each of which includes a modulator configured to each receive a stream partition and to generate modulated data from the received stream partition for transmission as two or more of the separate RF channels, a controller configured to generate signaling data and to control the two or more framing units to combine the signaling data with respect to the frames of data for transmission in the one of more separate RF channels, the signaling data being for configuring a receiver to detect and to recover the two or more stream partitions from the two or more RF channels and to recombine the two or more stream partitions into the data stream, wherein the signaling data for each of the RF channels includes an indication of bonding with at least one other RF channel and at least one radio frequency displacement between the radio frequency number of the RF channel and at least one other of the two or more separate RF channels.

2. A transmitter as claimed in Claim 1, wherein each of the two RF channels carries signaling data including the indication of bonding with at least one other RF channel and the radio frequency displacement to the other at least one RF channel.

3. A receiver for receiving data of a data stream via at least two separate RF channels, each of the RF channels carrying a stream partition of the data stream, the receiver comprising two or more radio frequency tuners, each of which is configured to detect and to isolate one of the two or more separate RF channels, two or more demodulators configured to each receive one of the two or more separate RF channels from one of the two or more radio frequency tuners and to demodulate the and the signaling and the data of a stream partition from the RF channel to form the stream partition, and a data stream combiner configured to combine the demodulated data of the two or more stream partitions from the two or more demodulators into the data stream, and a controller configured to recover the demodulated signaling data carried in one of the two or more separate RF channels, which provides an indication of bonding to at least one other RF channel and a radio frequency displacement which identifies a relative displacement of the at least one other RF channel to which the said RF channel is bonded and to control the two or more radio frequency tuners and the two or more demodulators to recover respective partition streams from the RF channels for the data stream combiner to recover the data stream.

4. A radio frequency translator comprising two or more radio frequency tuners, each of which is configured to detect and to isolate one of the two or more separate input RF channels, two or more amplitude modulators each of which is configured to generate the signal modulated on the input RF channel at a translated output frequency, and a controller configured to control the radio frequency tuners and the radio frequency amplitude modulators in response to configuration information input by a human, wherein the configuration information includes a radio frequency number of the other input RF channel to which this input RF channel is bonded and a radio frequency displacement which identifies a relative displacement of the at least one other RF channel used to control the two or more amplitude modulators to preserve the said configured radio frequency displacement in the translated output frequencies of the input and configured bonded input RF channels.

Intellectual

Property

Office

Application No: GB 1702012.4 Examiner: Daniel Voisey