<title>Management of heterogeneous traffic loading in DBS networks</title>

Abstract

In the paper we present the Adaptive Resource Allocation and Management (ARAM) algorithms developed to manage a Direct Broadcast Satellite (DBS) system supporting heterogeneous traffic mixes and operating under dynamic channel conditions. This traffic mix includes both: (i) data traffic that operates as an available bit rate flow and, (ii) video traffic that generates a variable bit rate flow. Both types of traffic use the Internet Protocol (IP) so they can be efficiently multiplexed on the same link. The dynamic channel conditions reflect time variation error rates due to external effects such as rain or jamming. ARAM attempts to maximize the utilization of the available capacity on the forward DBS link while maintaining Quality of Service (QoS) in the presence of congestion int he network and channel degradation effects. To achieve these ends, it utilizes adaptive control of video compression rates, data transmission rates, and channel forward error correction rates. One of the major features of ARAM is the admission control algorithm used to determine the number of variable bit rate flows admitted for service. In order to maximize the resource utilization, assignment of the variable bit rate services based on their peak rate is avoided. Instead, a flexible utilization of the bandwidth requiring the estimation of statistical multiplexing gain is used enabling more services to share the DBS link. Therefore in this paper, we focus on the ARAM admission control algorithm and assess its impact on QoS and DBS link utilization.