Fast and Robust Multi-class Signaling Overload Control Algorithms for 3G Wireless Switches

Mobile Switching Centers (MSCs) in wireless telecommunication networks must implement signaling overload controls to maintain acceptable performance during periods of high load especially due to fast changing hot spots resulting from user mobility. An MSC's signaling load comprises a variety of service requests, some more important than the others, and it must treat them differentially under overload conditions. We propose multi-class (to account for different types of requests) overload control algorithms that are highly reactive to sudden bursts of signaling load but also maintain high throughput and low average delay under a wide range of overload conditions. First, we introduce the concept of equivalent measures to convert the system load measures associated with different classes of traffic into a single measure associated with a pre-defined base class, thus allowing single-class overload measurement and detection algorithms to be used to monitor an equivalent traffic load. Next, we develop a new algorithm for partitioning the allowable equivalent load across traffic classes, using a strict priority scheme. Using simulations of call flows from the third generation (3G) Universal Mobile Telecommunications System (UMTS) standards, we compare different multi-class overload algorithms under a variety of overload conditions. Our simulation results indicate that our algorithm that measures system load using a combination of request acceptance rate and processor occupancy, provides highly reactive overload control. Last, for the purpose of making this algorithm more robust, we propose a measurement-based simple regression technique to dynamically estimate key system parameters. We find that estimates derived in this manner converge rapidly to their true values.