AU2020103224A4 - Secure Visible light communication through DRL-Smart Beamforming to protect against eavesdropping of wiretap - Google Patents

Abstract

Visible light communication is a smart communication technique for its unique features of license-free light spectrum, low cost of implementation and robustness against interference. The VLC provides line-of-sight transmission and better coverage property among its competitive communication techniques and are prone to security loopholes affecting actual legitimate users and network administrator. The visible light communication channel broadcasts faces the issue of VLC downlink susceptible to unauthorized terminals in offices and shopping zones. The transmitting information can be eavesdropped or wiretapped by malicious attackers in the light range. An efficient and secure transmission technique required to avoid eavesdropping. The scenario considered here is of multiple input single output (MISO) where there is multiple light fixtures to transmit and legitimate users and intruder who attacks the information through wiretap channel. The invention framework has smart beamforming over the MISO VLC wiretap channel which reduces the capability of information interference. The Deep reinforcement learning based VLC beamforming control scheme invention to handle dimensionality curse for both observation and action and to avoid quantization error in RL based algorithm. Fig 1 Streaming Control & I I Data Output Server Power Control Convertor - - I L1 L2 La %I % % LED lamps Internet r - - - - - - - - - - - -rV PD Detector ReceivedData Amplification & Receiver I Apply Beamforming Control results Measure previous policy HHSecrecy rate Fig 1DRL-Smart Beamforming Scenario

Description

Streaming Control I Data Output & I

Server Power Control Convertor

- - I L1 L2 La %I %

% LED lamps

Internet

r - - - - - - - - - - - - rV PD Detector

ReceivedData Amplification &

Receiver I

Apply Beamforming Control results Measure previous policy HHSecrecy rate

Fig 1DRL-Smart Beamforming Scenario

Editorial Note 2020103224 There is 7 pages of Description only.

Secure Visible light communication through DRL-Smart Beamforming to protect against eavesdropping of wiretap

Field and background of the invention

[001] Modern day communication requires safe and secure data transmission. The indoor data communication environment requires high rate transmission and vast coverage under limited spectrum of the conventional radio-frequency band, where the choice of communication is Visible Light Communication (VLC).in the VLC communication, there occur a threat of eavesdropping as the transmitter transmits data as multiple input and single output. The VLC communication broadcast is of Line-of-Sight (LoS) transmission and prone to security threats which can illegally retrieve data through wiretrapping. This becomes a barrier to VLC communication implementation and usage in many confidential data sharing. Security assuring mechanisms being readily applicable to such environments and at times are ineffective in many cases. Encrypted data been decrypted by potential hackers who could work on VLC transmission as legitimate user. Thus needed advanced technique and mechanism to control such eavesdropping in VLC transmission. The technique must get known of hacking attempts and derive a guideline to curtail the hacking possibilities. The hacker tries to wiretrap the legitimate channel through the unknown location where channel state information is unable to calculate. This finds the need for smart beamforming approach to work as Multiple Input Single Output (MISO) to strongly identify the user and admin and avoid eavesdropping. The potential hacker can hack such pocket of data if the sender and receiver information are known. The invention framework includes Visible Light Communication which provides secure data transfer through the Line-of-Sight (LoS) zone. The hacker may wiretap for retrieving the information and pose a great threat to secure data transfer.

Summary of the Invention

[002]The network helps to enable communication entities. The network started with message sharing to valid parties through the authenticated party identification like address, name and location. The modern day communication is improved as wireless network communication which has clients registered and given valid tokens to pass message and receive information. This way of communication has the problem of message security. The secure message transfer has been a challenge in any kind of data sharing. The security can be imposed by encrypting algorithms. The encrypting algorithms use the sender and receiver information as key to view the message. The potential hacker can hack such pocket of data if the sender and receiver information are known. The invention framework includes Visible Light Communication which provides secure data transfer through the Line-of-Sight (LoS) zone. The hacker may wiretap for retrieving the information and pose a great threat to secure data transfer. This can be prevented using deep-learning based visible beamforming approach which counts the eavesdropper attempts using calculative strategy and improves the receiver BER to transfer at the best possible secrecy assured. To adapt to the high-dimensional and consistent activity and state spaces all the more effectively and proficiently, a DRL-based MISO VLC beamforming control conspire has been acquainted with additional speed up and the learning execution of the keen beamforming based enemy of listening in framework. Reproduction results confirm that the proposed DRL-based plan can essentially build the mystery rate and utility, and reduction the BER of the authentic collector contrasted and the current benchmark conspire. Also, the proposed plans can approach or even beat the presentation of the zero-constraining beamformingThe channel state information of the hacker is known then the process of preventing data leak can be performed. Where the geographical location of the hacker keeps on changing and is unable to find, then the CSI has to be set as beamforming policy control readily available at the transmitter end has to made as zero-forcing beamforming thus to say no information to the eavesdropper. The invention includes smart transmitter identifies its beamforming policy based on the earlier security performance parameters like secrecy rate of the VLC system and the BER of the legitimate receiver which ensures zero transmission for illegal user. The DRL based smart beamforming method also impacts the learning rate and enhances performance by exploiting the complicated information and high-dimensional structure of beamforming policy domain which uses actor-critic approach and deep neural networks to study the network and decide the legitimate data transfer. This mechanism has to be improved with deep learning approach to make the receiver has maximum secrecy rate and has minimum data transfer hops. Brief description of the system

[003] Visible light communication (VLC) has been a next-generation wireless communication to transmit data in secure and efficient mechanism. The transmitter here is of light emitting diodes (LEDs) and receiver is of Photodiodes (PD). Consider an indoor MISO VLC scenario with LEDS fixed as down-facing light fixtures on the ceiling or the wall acts as transmitter to enable data transfer among legitimate users and prevent eavesdropping. The receiver is given photodiode to get data as the use of multiple receiver antennas may cost high and complex in installation. The LEDs are used more in number to transmit data and PDs are used less to ensure ease of data access and reduce complexity. The photodiodes convert light energy into electric energy.

[004]The secrecy rate should be high to ensure data transmission protectively. When the location of the eavesdropper is known then the channel state information can be derived and blocked from data transmission. The zero-forcing beamforming can be exploited to get optimal secrecy rate where the imposing zero beamforming and increasing more beamforming to actual legitimate user. The hacker's zero forming can be initiated by beamformer vector and VLC channel gain. The invention DRL-based beamforming framework with MISO VLC beamforming can perform better in continuous action space and high dimensional state space to prevent eavesdropping with higher security data transmission. The hacker may be located anywhere trying to hack the information through wiretap. The visible light communication here uses multiple transmitters which analyze the receiver receiving capacity with maximum secrecy and high BER. The invention deep learning based visible beamforming aporoach has the unique feature to set the beamforming vector for the hacker or the unauthorized attempt to tap the information is forced to zero. The enhanced simulations that the DRL-beamforming control method increases the utility and the secrecy rate of the MISO VLC system which reduces the BER of the legitimate reciver than the other commercial methods against eavesdropping.The zero-forcing method aim to find a beamforming vector nulled for eavesdropper and channel-access is restrained for the eavesdropper. This method does not improve the efficiency of the user channel BER by zero-forcing method. The DRL-based MISO VLC beamforming control scheme is required.

[005] When both legitimate user and eavesdropper are located in same location with a motive to receive data, then the secrecy rate has to be improved through the BER value of the user. The DRL-based smart beamforming method also impacts the learning rate and enhances performance by exploiting the complicated information and high-dimensional structure of beamforming policy domain which uses actor-critic approach and deep neural networks to study the network and decide the legitimate data transfer. The data available to the hacker has to be controlled using the smart beamforming approach as the eavesdropper has the receiving signal level set to '' using the beamforming control policy.

Description of the system

[006] The communication network poses a great challenge in data transfer as the level of secrecy between the transmitter and the receiver is always in stake. The visible light communication is an open end network with multiple users share the network. The transmitter used here is of the LEDs placed in the ceiling or the wall to enable communication, the receiver is of the PDs which convert light energy into electrical energy. This invention has the high performance exhibition of the zero-forcing strategy the arrangement space of the zero-driving technique is genuinely restricted and it can't locate the worldwide ideal arrangement of the framework. Contrasted and the zero-constraining strategy, the proposed RL based structure, particularly the DRL-VB calculation without quantization blunder, can outflank the zero-driving one and join to the ideal arrangement in the VLC hostile to listening in framework. The visible light communication here uses multiple transmitters which analyze the receiver receiving capacity with maximum secrecy and high BER. The information sharing through such open network are always an easy goose for the hackers to eavesdrop the information. The eavesdropped information can be of valued information. Thus any leak of data has to be arrested at the earliest. The hacker may be located anywhere trying to hack the information through wiretap. The receiver is given photodiode to get data as the use of multiple receiver antennas may cost high and complex in installation. The LEDs are used more in number to transmit data and PDs are used less to ensure ease of data access and reduce complexity. The photodiodes convert light energy into electric energy.

The visible light communication here uses multiple transmitters which analyze the receiver receiving capacity with maximum secrecy and high BER. The invention deep learning based visible beamforming aporoach has the unique feature to set the beamforming vector for the hacker or the unauthorized attempt to tap the information is forced to zero. The zero-forcing beamforming vector upon the unauthorized data access requester blocks the data transfer between the transmitter and the unauthorized user. The legitimate user who gets registered with the network has the advantageous feature of finding the best possible hops to reach the user. The lower and upper bound capacity with total average intensity constraint concluded for the constant parallel VLC channel assuming perfect CSI at the transmitter. The perfect secrecy capacatiy of the MIMO broadcast channel developed to guarantee the eavesdroppes receives no signal. The secrecy performance metric adopted to be of a precoding scheme for multi user MISO VLC broadcast channels. For secrecy protection, the beamforming technique is a key method in any network communication. The robust transmit beamformer is invented for the purpose of maximizing the achievable secrecy rate of MISO VLC System which are subject to amplitude modifications when the channel state information or location of the eavesdropper is known. An efficient and secure transmission technique required to avoid eavesdropping. The scenario considered here is of multiple input single output (MISO) where there is multiple light fixtures to transmit and legitimate users and intruder who attacks the information through wiretap channel. The invention framework has smart beamforming over the MISO VLC wiretap channel which reduces the capability of information interference. Another method excludes a MISO VLC wiretap situation has been considered, where a endeavors to wiretap the data which is initially sent for the genuine recipient. The invention of transmission attributes of the visible light communication (VLC) channel makes VLC downlinks defenseless to unapproved terminals in numerous real VLC situations, for example, workplaces and malls. This paper thinks about a multiple-input single-output (MISO) VLC situation with different light installations going about as the transmitter, a VLC recipient as the genuine client, and a snoop endeavoring to catch the undisclosed data. To improve the classification of VLC joins, a physical-layer against listening in system is proposed to cloud the unapproved busybodies and reduces their ability of inducing the data through keen beamforming over the MISO VLC wiretap channel. To adapt to the recalcitrant issue of finding the hypothetically ideal arrangement of the mystery rate and utility for the MISO VLC wiretapping channel, a reinforcement learning (RL) based VLC beamforming control conspire is proposed to accomplish the ideal beamforming strategy against the busybody. A learning-based enemy of snoopping structure by means of brilliant beamforming over the MISO VLC wiretap channel has been proposed to keep the busybody from wiretapping the mystery signals. To infer the ideal beamforming strategy, an RL-based MISO VLC beamforming control conspire has been intended for the MDP in a powerful climate the eavesdropper by artificial noise to reduce SNR over the MISO VLC channel wiretap. The mutual information and achievable secrecy rate for space shift keying is determined.

[007]The reinforcement learning process can set optimal beamforming policy through learning iterations read repeatedly with the prefixed time slots. The Deep-learning based smart beamforming framework uses Multiple Input Single Output VLC beamforming to analyze the user and the hacker through previous data transfer records and attempts. The invention DRL-based MISO VLC beamforming algorithm has continuous space action and high-dimensional state space to bring the efficient and effective anti-eavesdropping and high data security challenges. The utility and secrecy rate of the MISO VLC system when compared with RL-based scheme is very much appreciable in performance in DRL based VLC beamforming scheme. The Bit-Error-Rate (BER) of the legitimate user is also analyzed and shows a good sign of improvement to receive data without any delay or distortions.

[008] The invention smart transmitter determines the beamforming policy based on the security performance at the previous transfers like secrecy rate of the VLC system and the user proximity as well as the BER of the receiver. The beamforming policy control performed repeatedly to dynamically avoid eavesdropping communication. The q-learning based approach to RL has an effective beamforming selection for the MISO VLC system. The invention space is of continuous action space and high-diomensional where the q-learning has the limitation of discrete and low-dimensional action spaces efficiency. Thus the Deep Reinforcement learning based on VLC MISO beamforming is chosen and applied for the system.

Editorial Note 2020103224 There is 2 pages of Claims only.

Claims (6)

We Claim:

1. The visible light communication is most affordable and prompt for indoor communication in wireless technology. a. The data communication may rapidly change and decline depending upon the user strength and accessibility. b. The VLC beamforming control scheme enhances the observation and action space through deep reinforcement learning.

2. The Eavesdropping in any network is a major challenge for secure data transfer that is prevented and legitimate user data reception is enhanced. a. The network can be accessed by any user in open data communication network like Visible light communication. b. The VLC includes deep reinforcement learning to get knowledge of the earlier security hacks. c. The secrecy rate maximized by zero forcing beamforming over the VLC wiretap channel with multiple transmitters and single authorized receiver.

3. The Deep Reinforcement Learning based VLC scheme achieves the optimal beamforming policy. a. The smart beamforming policy prevent the hacker through previous attempt and define a adoptable approach to reduce eavesdropping wiretap attempt. b. The BER of the legitimate user is maximized through the algorithm to transmit data securely to the user. c. The photodiode here used as the receiver to receive the data transmitted in the VLC network.

4. The DRL-based smart beamforming policy increases the learning rate and performance by exploiting complex data and high-dimensional striucture of the beamforming policy through deep neural networks.

a. The DRL-based beamforming control method improves the utility and the secrecy rate of the MISO VLC system based on the invention methodology. b. The secrecy rate and the BER of the legitimate receiver to achieve the optimal beamforming policy. c. The DRL-based algorithm for MISO VLC smart beamforming control expoits actor-critic approach and deep Q network for learning and achieve anti-eavesdropping performance.

5. The Multiple-Input-Single-Output scenario helps to prevent wiretap channel in VLC framework. a. The multiple LEDs fixed in the transmission zone are used for identifying the hacker attempts in earlier transmission. b. The zero-forcing vector in beamforming ensures the hacker receives no signal during data transmission.

6. The DRL-based MISO VLC beamforming control scheme balances the legitimate user and the eavesdropper. a. The secrecy rate of the transmission is ensured through smart beamforming anfd improve desired signal intensity. b. The zero-forcing method finds the beamforming vector that the channel gain of identified hacker through deep reinforcement learning method is nulled.

Streaming Control 04 Nov 2020

Data Output

Server Power Control Convertor

L1 L2 Ln

LED lamps ..... 2020103224

Internet

PD Detector PD

Received Data Amplification & Processing Receiver

Apply Beamforming Control results Measure previous policy Secrecy rate

Fig 1DRL-Smart Beamforming Scenario