Tóm Tắt CRYPTO-CODING MECHANISM COMBINED WITH KEY AGREEMENT BASED ON THE CHANNEL CHARACTERISTICS FOR MULTI-ANTENNA WIRELESS COMMUNICATION SYSTEMS

The PLS methods that have been mentioned a lot in recent years are the secret key generation methods based on the wireless channel characteristics and the method of data encryption and e

MINISTRY OF EDUCATION AND TRAINING

HANOI UNIVERSITY OF SCIENCE AND TECHNOLOGY

DINH VAN LINH

CRYPTO-CODING MECHANISM COMBINED WITH KEY AGREEMENT BASED ON THE CHANNEL CHARACTERISTICS FOR MULTI- ANTENNA WIRELESS COMMUNICATION

The doctoral dissertation was accomplished in Hanoi University of Science and Technology

Supervisor: Prof Dr Vu Van Yem

At …… , date … month … year ………

The dissertation information can be found at following libraries:

1 Ta Quang Buu Library – Hanoi University of Science and Technology

2 National Library of Vietnam

1

INTRODUCTION Introduction

Nowadays, multi-antenna systems are widely deployed in modern wireless communication networks due to their outstanding advantages such as high data rates, enhanced communication efficiency, and better spectrum utilization These systems are essential for fulfilling the increasing need for seamless connectivity and enormous data transmission in various applications However, the broadcast nature of wireless communication makes these systems inherently vulnerable to security threats One of the most critical threats is the eavesdropping attack, in which adversaries can passively intercept confidential information by simply listening to the wireless channel Once this confidential information is captured, attackers may initiate a variety of follow-up attacks in an attempt to find the original content Therefore, it

is essential to enhance the security and reliability of multiple-antenna systems to ensure safe and trustworthy communication

Security for wireless communication systems mostly depends on traditional encryption methods at the upper layers However, the traditional encryption methods are computational security and their security levels are based on the difficult of the fundamental mathematical issues that they employ Moreover, the traditional encryption methods have some issues such as transmission delay, and timeout at higher layers As a result, physical layer security (PLS) is a potential technique to solve all the above problems The PLS methods that have been mentioned a lot in recent years are the secret key generation methods based on the wireless channel characteristics and the method of data encryption and error correction in the same step (crypto-coding technique) They can reduce the computational complexity and hardware structure, and they also ensure confidentiality and reliability

Motivation

The cryptosystems at high layers are employed to protect the secret

2

information Cryptosystems can be divided into two types such as symmetric-key encryption and asymmetric-key encryption Symmetric-key encryption is suitable for resource- limited devices However, the key distribution method is a major disadvantage of this system because a trusted party is required to manage and distribute the keys Moreover, when the secret key is stored in limited memory, the secret key can be reused acrosscommunication sessions, which reduces the confidentiality

of the secret keys Therefore, new key generation and distribution methods are being studied as alternatives to third-party key distribution methods Besides, the key generation technique based on the reciprocal characteristics of the wireless channel is a major topic in the field of PLS, which is the potential technique for key distribution and management Due to the transmission link in both transmitter and receiver being similar, the shared secret keys can be extracted the same for legitimate users The secret key can be changed by implementing channel estimation instead of depending on previous keys

So far, the majority of cryptographic methods have been designed independently with channel coding in secure wireless communication systems This increases complexity and latency in the classical systems

To address this issue, crypto- coding techniques have attracted a lot of attention from scientists This technique allows data encryption and error correction functions to be performed in the same step These crypto-coding techniques are based on Turbo codes, polar codes, and LDPC However, they have limitations in performance metrics, including error correction ability, computational complexity, and security, making them difficult to apply to advanced multi- antenna wireless communication systems

Objective, subject, and methodology research of the dissertation

3

 Proposing the crypto-coding techniques based on the Turbo codes and polar codes for the MIMO and massive MIMO systems, respectively

- Using the secret keys generated from the massive MIMO systems

to control the frozen bits and information bits of the polar codes

Conclusions and future works Conclusions with notable points

about the entire content, scientific contributions of the dissertation, as well as research directions in the future, will be given

4

CHAPTER 1 SECURITY AND RELIABILITY IN MULTI- ANTENNA WIRELESS COMMUNICATION SYSTEMS 1.1 Overview of multi-antenna wireless communication systems

The traditional multi-antenna communication systems are shown in Figure 1.1

Figure 1.1 The traditional multi-antenna communication systems

1.2 Encryption

1.2.1 Classification of cryptosystems

Symmetric-key encryption: A cryptosystem uses the same key in the encryption and decryption process Therefore the key must be kept secret

Asymmetric-key encryption: uses a public key for encryption and private key for decryption

5

systems are separated, which increases the computational complexity and delay To address these issues, this dissertation proposes two PLS approaches, including key generation based on wireless channel characteristics, and a joint data encryption and channel coding technique, referred to as the crypto-coding technique

1.2.4 Information-Theoretic basics for PLS

Figure 1.4 shows Shannon cipher system

Figure 1.4 Shannon cipher system

Wyner’s wiretap channel:

Figure 1.5 describes a scenario where a sender (Alice) wants to transmit secure data to a legitimate receiver (Bob) while an eavesdropper (Eve) can also receive an attenuated version of the signal Alice-Bob's transmission channel with good signal is considered as the main channel, while Eve also receives signal but with lower quality due

to signal degradation or noise is considered as the wiretap channel

Figure 1.5 Wyner’s wiretap channel model

A diverse array of PLS techniques has been explored in the literature, including power control-based approaches, chaos-based

Decoding process Original

data

Received data

6

methods, channel-aware schemes, lattice-based cryptographic systems, signal detection techniques, and theoretically secure capacity frameworks

1.3 Channel codes

1.3.1 Turbo codes

The diagram of the normal Turbo codes is displayed in Figure 1.6

Figure 0.1 The diagram of the normal Turbo encoder

1.3.2 Polar codes

A simple structure of polar codes can be shown in Figure 1.10

1.4 Wireless channel of multi-antenna communication systems Properties of wireless channel:

7

Spatial decorrelation

The multi-antenna model can be shown in Figure 1.12

Figure 1.12 The multi-antenna system model

1.4 Secret key generation based on wireless channel characteristics

Each step in detail is described in the subsections below

Step 1: Channel probing

Step 2: Quantization

Step 3: Information reconciliation

Step 4: Privacy amplification

8

channel with noise

Decryption in a step (Crypto-coding) Encryption in a step

(Crypto-coding)

Figure 1.17 The communication systems applying the crypto- coding

technique

Evaluation of security level for crypto-coding techniques

Bob’s BER and Eve’s BER will be applied to evaluate the security performance of the proposed crypto-coding techniques in Chapter 3

1.6 Conclusion

Chapter 1 provides a comprehensive overview of traditional antenna wireless communication systems, focusing on key components including encryption, channel coding, and wireless channel modeling Current systems often adopt conventional encryption and coding methods in a modular and independent manner While this approach ensures a degree of confidentiality and reliability, it introduces several challenges related to secret key management, increased system complexity, and transmission latency To resolve these limitations, two promising physical layer security (PLS) approaches have been introduced: secret key generation based on wireless channel characteristics, and crypto-coding techniques that integrate encryption and error correction in a unified process These methods not only aim to enhance security and reliability but also reduce computational overhead and latency In light of these considerations, this dissertation proposes novel solutions that combine crypto-coding techniques with secret keys derived from the physical properties of the wireless channel These proposed methods will be detailed in Chapters 2 and 3

multi-CHAPTER 2 SECRET KEY AGREEMENT METHODS BASED

ON WIRELESS CHARACTERISTICS FOR MULTIPLE ANTENNA COMMUNICATION SYSTEMS

2.1 Related works

Overall, the aforementioned research either did not evaluate the generated keys' randomness or the randomness degree is low As a result, in this dissertation, the key generation methods are proposed for the multi-antenna communication systems to increase the randomness

of the extracted secret keys The proposed methods will be presented in

Table 2.1 Simulation parameters of the MIMO systems

Alice Channel probing

Calculating the mean value of the real parts

Computing the

modulus of each

complex value of V i

Extracting real (Re i) and

imaginary (Im i) parts of

each complex value of V i

1A

K

2r K

1B K

value of real part

(Re i) with the mean value to obtain binary sequence   

Comparing each value of imaginary

part (Im i) with the mean value to obtain binary sequence

 

XOR operation 2im

K

2 A K

Bob Channel probing

Calculating the mean value of the real parts

Computing the modulus of each

complex value of V i

Extracting real (Re i) and

imaginary (Im i) parts of

each complex value of V i

Method 1 Method 2

Calculating the mean value of

W moduli

Comparing each

modulus Vi with the mean value to obtain binary key sequence

 w

   Calculating the mean value of the imaginary parts   

Comparing each value of real part

(Re i) with the mean value to obtain binary sequence

Comparing each value of imaginary

part (Im i) with the mean value to obtain binary sequence

 

XOR operation

2r K

2B K

Probing signal Coherence time The estimated

CoIR matrix

The estimated CoIR matrix Converting the CoIR matrix

into a single row vector (V)

Converting the CoIR matrix

into a single row vector (V)

Selecting W samples from

10

Simulation results of Method 1

For method 1, the MIMO systems are simulated with a BPSK modulation scheme in two case studies including AWGN and Rayleigh channels

Case study 1

The AWGN is simulated for the MIMO 2×2 system and the results are displayed in Table 2.3

Table 2.3 The results of method 1 for the case of the MIMO 2×2

system via AWGN channel

Case study 2

The MIMO systems are simulated through the Rayleigh channel The obtained p-values are shown in Table 2.5 for the MIMO 2×2

Table 2.5 The results of method 1 for the case of the MIMO 2×2 system

via the Rayleigh channel

Simulation results for method 2

There is a similarity to method 1, the MIMO systems are simulated through the AWGN and Rayleigh channels

Case study 1

11

The results of the AWGN channel are shown in Table 2.7

Table 2.7 The results of method 2 for the case of the MIMO 2×2

system via AWGN

Case study 2

The MIMO 2×2 system are simulated via the Rayleigh channel It can be observed from Table 2.9

Table 2.9 The results of method 2 for the case of the MIMO 2×2 system

via Rayleigh channel

Comparison of two methods

It can be seen that the method 1 outperforms the method 2 in terms of passing rate The average p-values of method 1 are greater than that of method 2 in most tests

2.3.2 Massive MIMO systems

In this section, the massive MIMO systems will be simulated via the Rician channel with different structures, including 128Tx-4Rx and 256Tx-16Rx

Simulation results for Method 1

12

Table 2.12 The results of method 1 for the case of massive MIMO

256Tx-16Rx system

Simulation results for Method 2

Table 2.14 The results of method 2 for the case of the massive MIMO

256Tx-16Rx system

Based on the simulation results and the comparison of the MIMO and massive MIMO systems, method 1 provides the greater passing rates and the average p-values Moreover, the calculations in method 1 are logical operations (XOR), so method 1 will require a simpler hardware configuration to execute Meanwhile, the process of generating a key sequence in method 2 involves algebraic operations and XOR operation, resulting in a higher computational complexity compared to method 1 Therefore, method 2 requires slightly more complex hardware and processing steps to execute Consequently, the key generation method based on the modulus of the CIR is suggested to extract the secret key for the MIMO and massive MIMO systems

2.4 Summary

Chapter 2 depicts the key generation methods based on the wireless properties Through analyzing the wireless channel properties, two CIR-based methods are proposed for the MIMO and massive MIMO systems After evaluating the performance of the proposed methods, the following points can be noticed:

13

- The CIR of the multi-antenna communication systems is the infinite source of randomness that can be used to generate the secret key

- The simple quantization algorithms are applied to convert the CIR components into the secret keys

- The key length can be readily altered while maintaining the randomness degree of secret keys

The research results on key generation methods in Chapter 2 are published in publications 2 and 4 of publication lists

CHAPTER 3 CRYPTO-CODING TECHNIQUES FOR THE MULTI-ANTENNA COMMUNICATION SYSTEMS

3.1 Related works

The previous crypto-coding based on Turbo codes have not yet achieved good performances in terms of security level, error correction ability, and computational complexity They often trade off one of the three performance metrics to gain another Moreover, these methods still use traditional key distribution and management To be compatible with current advanced wireless communication systems, the secure Turbo methods must be implemented for MIMO systems that ensure standards such as low complexity, high error correction efficiency, and effectiveness against eavesdropping attacks

Most studies for polar codes have not fully evaluated performance in terms of computational complexity, error correction, and confidentiality They are mostly evaluated on a single antenna system via the Additive white Gaussian noise (AWGN) channel Therefore, they are not suitable for practical applications for 5G systems that use multi-antenna systems via the Rician channel

3.2 The multi-antenna systems using crypto-coding methods

3.3 Crypto-coding techniques based on Turbo codes 3.3.1 The proposed secure Turbo code by controlling the puncturing block (method 1)

3.3.1.1 Diagram of method 1

The proposed method for controlling the puncturing block is shown in Figure 3.2 and it will be described as follows