Optimal Dynamic Resource Allocation for Multi-Antenna Broadcasting with Heterogeneous Delay-Constrained Traffic

TL;DR

This paper develops an optimal dynamic resource allocation strategy for multi-antenna wireless broadcast channels supporting heterogeneous delay-sensitive and delay-tolerant data, balancing capacity and delay constraints with an efficient online algorithm.

Contribution

It introduces a convex optimization-based online resource allocation method for multi-antenna broadcast channels with mixed delay requirements, and analyzes the fundamental throughput-delay tradeoff.

Findings

Proposed an efficient cross-layer resource allocation algorithm.

Characterized the capacity loss due to delay constraints.

Analyzed the throughput penalty for zero-delay transmission.

Abstract

This paper is concerned with dynamic resource allocation in a cellular wireless network with slow fading for support of data traffic having heterogeneous transmission delay requirements. The multiple-input single-output (MISO) fading broadcast channel (BC) is of interest where the base station (BS) employs multiple transmit antennas to realize simultaneous downlink transmission at the same frequency to multiple mobile users each having a single receive antenna. An information-theoretic approach is taken for characterizing capacity limits of the fading MISO-BC under various transmission delay considerations. First, this paper studies transmit optimization at the BS when some users have delay-tolerant ``packet'' data and the others have delay-sensitive ``circuit'' data for transmission at the same time. Based on the convex optimization framework, an online resource allocation algorithm is derived that is amenable to efficient cross-layer implementation of both physical (PHY) -layer multi-antenna transmission and media-access-control (MAC) -layer multiuser rate scheduling. Secondly, this paper investigates the fundamental throughput-delay tradeoff for transmission over the fading MISO-BC. By comparing the network throughput under completely relaxed versus strictly zero transmission delay constraint, this paper characterizes the limiting loss in sum capacity due to the vanishing delay tolerance, termed the delay penalty, under some prescribed user fairness for transmit rate allocation.