Controlled Barrage Regions: Stochastic Modeling, Analysis, and Optimization

TL;DR

This paper develops a stochastic model for Controlled Barrage Regions in broadcast relay networks, analyzing their dynamics, outage probability, and optimizing network parameters for maximum capacity.

Contribution

It introduces a Markov process model for packet transmission in CBRs, deriving closed-form outage probabilities and optimizing network parameters.

Findings

Closed-form outage probability expressions accounting for fading and interference.

A Viterbi-like algorithm for modeling temporal correlation of active packets.

Optimized network parameters for maximum transport capacity.

Abstract

A barrage relay network (BRN) is a broadcast oriented ad hoc network involving autonomous cooperative communication, a slotted time-division frame format, and a coarse slot-level synchronization. While inherently a broadcast protocol, BRNs can support unicast transmission by superimposing a plurality of controlled barrage regions (CBRs) onto the network. Within each CBRs, a new packet is injected by the unicast source during the first time slot of each new radio frame. When a CBRs is sufficiently long that a packet might not be able to reach the other end within a radio frame, multiple packets can be active at the same time via spatial pipelining, resulting in interference within the CBRs. In this paper, the dynamics of packet transmission within a CBRs is described as a Markov process, and the outage probability of each link within the CBRs is evaluated in closed form, thereby accounting for fading and co-channel interference. In order to account for the linkage between simultaneous active packets and their temporal correlation, a Viterbi-like algorithm is used. Using this accurate analytical framework, a line network is optimized, which identifies the code rate, the number of relays, and the length of a radio frame that maximizes the transport capacity.