Predicting resource demand in heterogeneous active networks

Abstract

Recent research, such as the Active Virtual Network Management Prediction (AVNMP) system, aims to use simulation models running ahead of real time to predict resource demand among network nodes. If accurate, such predictions can be used to allocate network capacity and to estimate quality of service. Future deployment of active-network technology promises to complicate prediction algorithms because each "active" message can convey its own processing logic, which introduces variable demand for processor (CPU) cycles. This paper describes a means to augment AVNMP, which predicts message load among active-network nodes, with adaptive models that can predict the CPU time required for each "active" message at any activenetwork node. Typical CPU models cannot adapt to heterogeneity among nodes. This paper shows improvement in AVNMP performance when adaptive CPU models replace more traditional non-adaptive CPU models. Incorporating adaptive CPU models can enable AVNMP to predict active-network resource usage farther into the future, and lowers prediction overhead. INTRODUCTIONGrowing availability of processing power and bandwidth in communication networks encourages innovative approaches to network management. One specific innovative idea envisions injecting simulation models into network nodes, and then running those models in parallel with the operational network, but ahead in time, in order to predict traffic and resource use. If the models predict accurately, then network management systems can better allocate capacity in anticipation of varying demands and network operators can better estimate the quality of service (QoS) that customers can expect. This paper describes one approach, the Active Virtual Network Management Prediction (AVNMP) system [1], which aims to predict network traffic. AVNMP uses active-network technology [2] to inject simulation models into network nodes, and to run those models concurrently with corresponding applications. AVNMP then compares estimated performance against measured

Improving Architecture-Based Self-Adaptation through Resource Prediction

“…The Active Virtual Network Management Prediction System uses simulation models running ahead of real time to predict resource demand among network nodes. Such predictions can be used to allocate network capacity in anticipation of demand increase, and to ensure adequate quality of service to different network flows [9]. Our work shares theoretical foundations with these, but the problem domains are different.…”

Smart Messages: A Distributed Computing Platform for Networks of Embedded Systems

“…Each SM migrates through the network until it arrives at a node already visited by another discovery SM (i.e., it ends its execution) or it reaches the maximum number of hops that it is allowed to migrate. Once this threshold is reached, the SM migrates back one hop and reports its current information (lines [19][20][21][22][23][24][25][26][27][28]. This is a recursive process that Figure 17.…”

“…A significant amount of research has been done to solve this problem for real time systems [52,55] and active networks [21,40]. Given that we did not want to limit the expressibility of the programming language (e.g., SNAP [40]), our solution is based on user-provided lower bounds for resources and non-preemptive execution.…”

Gradient method with topology discovery for load-balancing in active networks

“…Active Networks [1] are programmable networks where packets can carry executable code, or reference to executable code [2]. The code is being run at execution environment (further referenced as EE) implementing active networking protocol specific processing.…”

“…Also, the process assignment algorithm and its parameters can be refined during the application's runtime or several versions may be running in the network. Another possibility is to use the applications like AVNMP (Active Virtual Network Management Prediction) [2,8,9] or to modify network's behavior to process's favor. These considerations lead us to the idea that for environments as Active Networks are, new methods for load balancing are needed.…”