1: \begin{abstract}

2:   Network-coded multiple access (NCMA) is a communication scheme for

3:   wireless multiple-access networks where physical-layer network

4:   coding (PNC) is employed. In NCMA, a user encodes and spreads its

5:   message into multiple packets. Time is slotted and multiple users

6:   transmit packets (one packet each) simultaneously in each

7:   timeslot. A sink node aims to decode the messages of all the users

8:   from the sequence of receptions over successive timeslots.  For

9:   each timeslot, the NCMA receiver recovers multiple linear

10:   combinations of the packets transmitted in that timeslot, forming a

11:   system of linear equations. Different systems of linear equations

12:   are recovered in different timeslots.  A message decoder then

13:   recovers the original messages of all the users by jointly solving

14:   multiple systems of linear equations obtained over different

15:   timeslots. We propose a low-complexity digital fountain approach for

16:   this coding problem, where each source node encodes its message into

17:   a sequence of packets using a fountain code. The aforementioned

18:   systems of linear equations recovered by the NCMA receiver

19:   effectively couple these fountain codes together.  We refer to the

20:   coupling of the fountain codes as a linearly-coupled (LC)

21:   fountain code. The ordinary belief propagation (BP) decoding

22:   algorithm for conventional fountain codes is not optimal for LC

23:   fountain codes. We propose a batched BP decoding algorithm and

24:   analyze the convergence of the algorithm for general LC fountain

25:   codes.  We demonstrate how to optimize the degree distributions and

26:   show by numerical results that the achievable rate region is nearly

27:   optimal. Our approach significantly reduces the decoding complexity

28:   compared with the previous NCMA schemes based on Reed-Solomon codes

29:   and random linear codes, and hence has the potential to increase

30:   throughput and decrease delay in computation-limited NCMA systems.

31: \end{abstract}

32: