Throughput increase of ALOHA-based systems with multiuser sequence detection

Abstract

Multi-user detectors have been suggested to improve the performance of CDMA based random-access systems [1, 2, 3, 4]. It is shown that with the maximum-likelihood (ML) multi-user sequence detector (MUSD) [1], spreading codes are often unnecessary in the AWGN channel; the MUSD can separate interfering packets of data using the same symbol pulse. A bit error probability of 10⁻⁴ is attained at signal-to-noise ratios of 12 to 30 dB provided the packets are offset slightly. The offset may be considered as a probable outcome in many systems, with failure (collision) occurring when the offset is too small. Alternatively, the desired offset may be designed into the system in a more controlled way. A generalization of ALOHA is given where ‘slot’-collisions are replaced by ‘subslot’-collisions. Throughputs obtained depend on the complexity available. For example, relative to a single user continuously transmitting, 91% throughput is achieved with 4 subslots, which requires an 8-state VerdÚ-Viterbi trellis detection algorithm, while 159% throughput is achieved with 10 subslots, requiring 512 states. Potentially, systems with K-subslots obtain K carrier-sense channels when carrier-sensing is available for each subslot.

This work is based on writings found in the Ph.D. thesis of the first author