Asynchronous Transmission for Multiple Access Channels: Rate-Region Analysis and System Design for Uplink NOMA

Abstract

In this work, we thoroughly analyze the rate-region provided by the asynchronous transmission in multiple access channels (MACs). We derive the corresponding capacity-regions, applicable to a wide range of pulse shaping methods. We analytically prove that asynchronous transmission enlarges the capacity-region of MACs. Although successive interference cancellation (SIC) can achieve the optimal sum-rate for the conventional uplink non-orthogonal multiple access (NOMA) methods, it is unable to achieve the boundary of the capacity-region for the asynchronous transmission. We demonstrate that for the asynchronous transmission, the optimal SIC decoding order to achieve the maximum sum-rate is based on the users' channel strengths. This optimal ordering is in contrast to the conventional uplink NOMA, where various decoding orders can result in the maximum sum-rate. Furthermore, we provide practical transceiver designs to approach the capacity-region. The memory induced by asynchronous transmission enables the use of the trellis-based detection methods which improves the performance. In addition, we propose a transceiver design, based on channel diagonalization to exploit the frequencyselectivity introduced by timing offsets. The proposed transceiver design, joint with the turbo principle, enables us to achieve a rate pair that is not achievable by the synchronous transmission.

Capacity Region of Asynchronous Multiple Access Channels with FTN

“…. , K and obtain K sets of samples instead of a single set [7]. Then, the samples y l [n] corresponding to user l are written by sampling the output of the matched filter, y(t), at time nδT + τ l , n = 0, .…”

“…A realistic problem follows, in practice, that each device will experience a random time delay. However, instead of being a hazard to the system, this asynchronism is analyzed in [6], [7] and [8] and is shown to be beneficial to multiple access transmission. In [6], the author explored the capacity region of asynchronous MAC with fixed or random time delay differences and showed that these differences bring in additional gains.…”

“…However, the analysis in [6] is constrained to rectangular pulse shapes in time. The authors of [7] extended this limitation and derived the capacity region for band-limited pulse shapes. In [8], the authors studied FTN in asynchronous MAC and obtained an achievable rate region with fixed power allocation.…”

Low-Complexity Dynamic Resource Scheduling for Downlink MC-NOMA Over Fading Channels

“…With the performance metric function in (6), the QoS constraints in (7), and the feasible sets for decision variables in (8) problem for joint user, subchannel, and power scheduling in the downlink MC-NOMA system over time-varying fading channels as…”

“…Unlike OMA techniques without inter-user interference (IUI), NOMA allows IUI by multiplexing multiple users on the same resource based on superposition coding in the power domain. Then, at the receiver, multi-user detection is realized by mitigating IUI based on successive interference cancellation (SIC) (see [7], [8] and references therein).…”