Concurrency and Computation: Practice and Experience Theoretical Computer Science

Research and Simulation Analysis of Multi‐Peer Parallel Services in Hybrid P2P Networks With Malicious Disturbances and Energy‐Saving Switching

Jinying Zhang, Zhanyou Ma
Wiley · 2026-11-01 · DOI 10.1002/cpe.71011

Abstract

ABSTRACT This article investigates the resource distribution and request processing problem in hybrid peer‐to‐peer (P2P) networks, focusing on the combined effects of malicious disturbances, energy‐saving switching mechanisms, and service‐time variability in a multi‐peer parallel service environment. In the considered system, resource peers with stable online availability and strong uploading capabilities constitute a service peer cluster, while request peers continuously generate resource acquisition tasks. Connection interruptions, request cancellations, timeout failures and malicious behaviors are uniformly modeled as disruptive disturbances to normal service requests. To characterize the dynamic switching process of the service peer cluster between the normal service mode and the low‐power lightly loaded mode, energy‐saving operation mechanisms, natural recovery mechanisms and threshold‐triggered switching strategies are introduced. Meanwhile, a queueing model with highly variable service times is established. Based on the Counting Service Phase Process (CSFP), the system state space and the corresponding continuous‐time Markov chain are constructed. The steady‐state distribution is derived, and analytical expressions for several important performance measures are obtained, including the mean system size, mean waiting time, mean sojourn time, effective throughput and energy consumption. Numerical results demonstrate that the request arrival rate, disturbance intensity, recovery parameters, threshold policies jointly affect system congestion, service reliability and energy consumption per successful request. Furthermore, benchmark comparisons, ablation experiments, robustness analyses, and discrete‐event simulations verify the effectiveness of the proposed model and demonstrate its capability to characterize the performance variations of hybrid P2P resource service systems under different operating conditions.

Read at publisher →