所謂的交換處理系統(Switched Processing Systems; SPS)被廣泛的應用在許多領域上，如通訊、電腦或製造業網路。這些系統通常包含不同的服務模組，其特徵是具有彈性(或可分享)且彼此之間不獨立。以往皆發展使吞吐量最大的控制策略，但近年則開始著重改善其服務品質(QoS)。本文將提出一個以抽樣及測量為基準的控制策略，此控制策略有以下之優點：(i)可大量節省資料儲存空間；(ii)可監控、預測及反應輸入流量之變動；(iii)可減少服務模組的交換次數(成本)；(iv)可達到系統的最大吞吐量(throughput)；並(v)可改善系統其他表現值(如工作的等待時間等)。最後，我們透過電腦模擬的方式來評估所提控制策略在各方面的表現。

Switched Processing Systems (SPS) represent a canonical model for many areas of applications of communication, computer and manufactur- ing systems. They are characterized by flexible, interdependent service capabilities and multiple classes of job traffic flows. Recently, increased attention has been paid to the issue of improving quality of service (QoS) performance in terms of delays and backlogs of the associated scheduling policies, rather than simply maximizing the system’s throughput. In this study, we investigate a sampling and measurement based control policy for switched processing systems that achieves the following objectives: (i) save a lot of memory, (ii) prediction and monitoring for intensity, (iii) re- duce the switch times of service modes, (iv) achieve maximal throughput in the systems and (v) improves performance with respect to delay. Finally, we perform the simulation study to show the performance assessment using propose policy.

[1] M. Armony and N. Bambos, “Queueing Dynamics and Maximal Throughput Scheduling in Switched Processing Systems” , Queueing Systems: Theory and Applications, 44(3), pp. 209-252 , 2003.
[2] N. Bambos and G. Michailidis, “Queueing Networks in Random Environments”, Advances in Applied Probability, 36, pp. 293-337, 2004.
[3] N. Bambos and G. Michailidis, “Queueing Networks of Random Link Topology: Stationary Dynamics of Maximal Throughput Schedules”, Queueing Systems, 50, pp. 5-52, 2005.
[4] J.G. Dai and B. Prabhakar, “The Throughput of Data Switches with and without Speedup”, Proceedings of IEEE INFOCOM, pp.556-564, 2000.
[5] Y.C. Hung, “Modeling and Analysis of Stochastic Networks with Shared Resources”, Ph.D.thesis, Department of Statistics, The University of Michigan, 2002.
[6] G. Michailidis, “Optimal Resource Allocation in a Queueing System with Shared Resources”, Proceedings of the 42nd Conference of Decision and Control, 2003.
[7] L. Tassiulas and P.P. Bhattacharya, “Allocation of Interdependent Resources for Maximal Throughput”, Stochastic Models, 16(1), pp. 27-48, 1999.
[8] L.Tassiulas and A. Ephremides, “Stability properties of constrained queueing systems and scheduling for maximum throughput in multihop radio networks”, IEEE Transactions on Automatic Control, 37(12), pp. 1936-1949, 1992.
[9] K.M. Wasserman, G. Michailidis and N. Bambos, “Optimal Processor Allocation to Differentiated Job Flows”, Performance Evaluation, 63, pp. 1-14, 2006.
[10] Y.C. Hung and G. Michailidis. “A Measurement Based Dynamic policy for Switched Processing Systems”, Proceedings of IEEE International Conference on Communications, 2007.
[11] Y.C. Hung and G. Michailidis. “Improving Quality of Service for Switched Processing Systems”, Proceedings of 11th International Workshop on Computer-Aided Modeling , Analysis and Design of Communication Links and Networks, 2006.
[12] K. Ross and N. Bambos, “Dynamic Quality of Service Control in Packet Switch Scheduling”, Proceedings of IEEE International Conference on Communications, 2005.
[13] K. Ross and N. Bambos, “Optimizing Quality of Service in Packet Switch Scheduling”, Proceedings of IEEE International Conference of Communications, 2004.
[14] Nick Duffield, “Sampling for Passive Internet Measurement : A Review”, Statistical Science, 19(3), pp.472-498, 2004.
[15] C.A. Lowry and W.H. Woodall, “A Multivariate Exponentially Weighted Moving Average Control Chart”, Technometrics, 34(1), pp.46-53, 1992.
[16] National Laboratory for Advanced Network Research, http://moat.nlanr.net/Images/Nvabar/2_MNA.html.
[17] Lili Yang, G. Michailidis, “Sampled Based Estimation of Network Traffic Flow Characteristics”, INFOCOM 2007, 26th IEEE International Conference on Computer Communications, pp. 1775-1783, 2007