Throughput Calculations for Multihop Packet Radio Networks

The main result of this paper is an efficient algorithm for evaluating throughput in a multihop packet radio network (PRN). The algorithm is needed for a Differentiated Services Quality of Service (QoS) framework for managing the allocation and policing of throughput in a PRN. But the known algorithm for conducting analyzing throughput in a PRN is not scalable because it requires computation on the order of an exponential of the size of the array, and it requires complete knowledge of the network topology. This paper proposes and evaluates an approximation to the original algorithm which scales linearly with network size and requires only local network topology. INTRODUCTION The Differentiated Services architecture for network QoS is being developed within an IETF (Internet Engineering Task Force) working group [1]. The key idea of the architecture is to minimize the processing and storage at the core routers which make up most of the network, by carefully controlling traffic as it enters the network. Nodes at entry points perform classification, policing, and shaping to force compliance to agreed flow rates. Compliant traffic is marked by setting bits in the IP header. Core routers inspect these bits when scheduling packets for forwarding. Various services can be built up using these principles. For example, a high priority, low delay service can be implemented by arranging for high priority traffic to receive preferred service compared to the rest of the traffic which is handled as besteffort. Making sure that the priority traffic actually receives low delay through the network requires a throughput model of the network, so that the service can make rational allocation decisions, a process known as provisioning. Based on these concepts, we are developing a QoS framePrepared through collaborative participation in the Advanced Telecommunications & Information Distribution Research Program (ATIRP) Consortium sponsored by the U.S. Army Research Laboratory under the Federated Laboratory Program, Cooperative Agreement DAAL01-96-2-0002. The U.S. Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation thereon. work for multihop packet radio network of mobile stations. Our framework has the following features: 1. A local QoS Manager at all stations to interact with applications and implement the distributed functionality of the service. 2. Applications connect with the QoS Manager and request high priority throughput to a selected destination. 3. QoS Managers aggregate QoS requests to estimate demand for high priority link flow. 4. QoS Managers estimate available throughput for point to point flows from the aggregated requests. 5. QoS Managers adjust allocations of throughput by fairly scaling demand. These allocations change over time in response to changes in policy, requests, and network state (topology and/or routing). 6. The QoS Manager delivers notifications to applications of the actual QoS allocations (throughput) and reallocations over time. 7. QoS Managers setup and manage the priority service within the network layer, by setting up priority queues, and arranging for high priority socket stream packets to be tagged and subsequently placed in a higher priority queue by the network layer. 8. QoS Managers arrange for the active shaping or policing of priority traffic to comply with allocations. (Shaping forces compliance by delaying out-of-bound packets. Policing leaves non-compliant packets unmarked.) The focus of this paper is the algorithm for estimating throughput in multihop packet radio networks, which is a key requirement of this framework. As discussed in an overview article by Tobagi [3], modeling packet radio networks is a more formidable task compared to point-to-point networks. The key sticking point is the channel. In point-topoint networks, the channel is dedicated to a single pair of nodes. In a PRN, the channel is a multi-access broadcast resource. In this regard, the channel of a PRN operates much like an Ethernet, where all nodes on the Ethernet contend for bandwidth. Yet in a PRN, each node is a member of a complex topology of Ethernet-like domains.