Abstract

Abstract-We present a comprehensive delay performance measurement and analysis in an operational large-scale urban wireless sensor network. We build a light-weight delay measurement system in such a network and present a robust method to calculate per-packet delay. . Those models and analysis assume specific network conditions, e.g., heavy traffic and fixed forwarding path, which are not always satisfied in real WSNs. Moreover, those studies lack comparison and validation with a real-world large-scale network. There are also many research works on delay analysis and measurement in Internet and data centers. Kompella et al. [16] present a fine-grained latency measurement method in presence of packet losses for Internet with a lossy difference aggregator. This method can measure per-packet delay in Internet while incurring very limited additional traffic overhead. To measure the per-flow delay, Lee et al. [17] present a measurement method with reference delay interpolation. This work extends existing works to efficiently measure per-flow delay, which is important for QoS in real applications. As the development of data center technologies, Wilson et al. [18]