Joint network and fountain codes design for relay-assisted multi-user system

Abstract

96wireless systems such as IMT-Advanced, a relay station needs to assist uplink transmissions originated from multiple user equipment within its coverage simultaneously [1] .General speaking, network coding (NC) is always used in efficient transmission. Full diversity can be achieved in NC cooperation system [2] . Rateless codes(RC), also known as fountain codes, are good at channel adaptability. The cooperation scheme based on RC has advantage on robustness [3] , throughput [4] , computation complexity, memory usage and overall energy consumption [5] . Well then, based on the above-mentioned two kinds of codes, joint network and rateless coding scheme is more preponderant for transmission efficiency, robustness and reliability. Ernest Kurniawan etc. [6] have proposed a novel scheme, which incorporated NC into rateless transmission. Partial information about the upcoming message block is allowed to be transmitted during the current block transmission, which reduces the required number of transmissions and improves the spectral efficiency. Then, Andrew Liau etc. [7] have proposed a new protocol, namely soliton-like rateless coding(SLRC), by exploiting the benefits of fountain coding and NC coding over a Y-network. SLRC exhibits a 5% reduction in overhead compared to the state of the art distributed LT(DLT) codes at high decoding success rates. Furthermore, Kun Pang etc. [8] have proposed a distributed Abstract: In relay-assisted multi-user system, relay coding is important to enhance the robustness and reliability of cooperative transmission. For better adaptability and efficiency, two joint network and fountain coding (JNFC) schemes are proposed. When the condition of all direct channels is worse, JNFC scheme based on distributed LT(DLT) codes is used. Otherwise, JNFC scheme based on multi-dimensional LT(MD-LT) codes is suited. For both two above-mentioned schemes, the united degree distribution design method for short-length fountain codes is proposed. For the latter scheme, MD-LT codes are proposed for equal error protection (EEP) of each user. Simulation results and analysis show that the united degree distribution need less decoding overhead compared with other degree distribution for short-length fountain codes. And then, all users are protected equally in despite of asymmetric uplinks.

Encryption scheme against eavesdropper collusion in wireless multi‐user network

“…It is assumed that the frame length of user is 500. Based on the degree distribution method in [33], the degree distribution at user node is Ω U ( x ) = 0.0252 x + 0.4829 x 2 + 0.1571 x 3 + 0.1057 x 4 + 0.0523 x 6 + 0.0664 x 7 + 0.0384 x 4 + 0.031 x 15 + 0.007 x 46 + 0.034 x 47 . We assumed that M correct = 4, the degree distribution is Ω R ( x ) = 0.3704 x 2 + 0.4224 x 3 + 0.0124 x 7 + 0.1177 x 8 + 0.0546 x 15 + 0.0102 x 89 + 0.0123 x 90 .…”

“…According to different channel condition, the degree distribution is also different. For multi‐user system with relay, the degree distribution design is shown in [33].…”

Concatenated physical layer encryption scheme based on rateless codes

“…The index sequence generation algorithm is as follows. Algorithm 1 Index sequence generation (1) Partition [0, 1] into L intervals uniformly, and the index of intervals is denoted as (1, 2, …, L ). (2) Using key as a seed to generate a random number a in [0, 1], which obeys uniform distribution. (3) Find the interval which a is in, and output the index of the interval as . Step 2 : Generate random degree sequence based on the degree distribution (the method can be seen in [39]) and where . The degree sequence generation algorithm can be seen as below: Algorithm 2 Degree sequence generation (1) Partition [0, 1] into (the maximum number of degree) intervals, the width of each interval is equal to the probability of each degree in . (2) Using key as a seed to generate a random number a in [0, 1], which obeys uniform distribution. (3) Find the interval which a is in, and output the corresponding degree . Step 3 : Select values from I in sequence to form the index sequence.…”

“…On the basis of the degree distribution design method in [39], distributions with length and are shown as On the basis of Fig. 4, RS(255,239) is assumed as the channel code in the simulation system.…”

Goal‐oriented design of optimal degree distribution for LT codes

“…Later, raptor codes were proposed to provide better performance by precoding [3]. Following the advent of rateless codes, LT codes have been introduced and exploited in both data transmission [4] and storage [5]. Generally, given k input symbols, potentially infinite number of encoded symbols can be generated and then transmitted to the receiver on the fly.…”