Iraj Saniee

Joint Scheduling and Congestion Control in Mobile Ad-Hoc Networks

In this paper we study the problem of jointly performing scheduling and congestion control in mobile ad-hoc networks so that network queues remain bounded and the resulting flow rates satisfy an associated network utility maximization problem. In recent years a number of papers have presented theoretical solutions to this problem that are based on combining differential-backlog scheduling algorithms with utility-based congestion control. However, this work typically does not address a number of issues such as how signaling should be performed and how the new algorithms interact with other wireless protocols. In this paper we address such issues. In particular: ldr We define a specific network utility maximization problem that we believe is appropriate for mobile adhoc networks. ldr We describe a wireless greedy primal dual (wGPD) algorithm for combined congestion control and scheduling that aims to solve this problem. ldr We show how the wGPD algorithm and its associated signaling can be implemented in practice with minimal disruption to existing wireless protocols. ldr We show via OPNET simulation that wGPD significantly outperforms standard protocols such as 802.11 operating in conjunction with TCP. This work was supported by the DARPA CBMANET program.

Light core and intelligent edge for a flexible, thin-layered, and cost-effective optical transport network

We present a new optics-based transport architecture that emulates fast switching in the network core via emerging fast tunable lasers at the network edge, and bypasses the need for fast optical switching and buffering. The new architecture is capable of handling both asynchronous and synchronous traffic, for dealing with various bandwidth granularities and responding to dynamic changes in end-to-end traffic demands. The architecture also reduces the amount of layering in the transport network by eliminating packet and TDM switching, keeps the network core light (lightweight and transparent), and pushes intelligence to the network edge. We discuss technical challenges that arise in the new architecture and describe possible approaches to address them.

Performance impacts of multi-scaling in wide area TCP/IP traffic

Recent measurement and simulation studies have revealed that wide area network traffic has complex statistical, possibly multifractal, characteristics on short timescales, and is self-similar on long timescales. In this paper, using measured TCP traces and queueing simulations, we show that the fine timescale features can affect performance substantially at low and intermediate utilizations, while the coarse timescale self-similarity is important at intermediate and high utilizations. We outline an analytical method for estimating performance for traffic that is self-similar on coarse timescales and multi-fractal on fine timescales, and show that the engineering problem of setting safe operating points for planning or admission control can be significantly affected by fine timescale fluctuations in network traffic.

Large-scale curvature of networks.

Understanding key structural properties of large-scale networks is crucial for analyzing and optimizing their performance and improving their reliability and security. Here, through an analysis of a collection of data networks across the globe as measured and documented by previous researchers, we show that communications networks at the Internet protocol (IP) layer possess global negative curvature. We show that negative curvature is independent of previously studied network properties, and that it has a major impact on core congestion: the load at the core of a finite negatively curved network with N nodes scales as N(2), as compared to N(1.5) for a generic finite flat network.

Scheduling bursts in time-domain wavelength interleaved networks

A time-domain wavelength interleaved network (TWIN) (Widjaja, I. et al., IEEE Commun. Mag., vol.41, 2003) is an optical network with an ultrafast tunable laser and a fixed receiver at each node. We consider the problem of scheduling bursts of data in a TWIN. Due to the high data rates employed on the optical links, the burst transmissions typically last for very short times compared with the round trip propagation times between source-destination pairs. A good schedule should ensure that: 1) there are no transmit/receive conflicts; 2) propagation delays are observed; 3) throughput is maximized (schedule length is minimized). We formulate the scheduling problem with periodic demand as a generalization of the well-known crossbar switch scheduling. We prove that even in the presence of propagation delays, there exist a class of computationally viable scheduling algorithms which asymptotically achieve the maximum throughput obtainable without propagation delays. We also show that any schedule can be rearranged to achieve a factor-two approximation of the maximum throughput even without asymptotic limits. However, the delay/throughput performance of these schedules is limited in practice. We consequently propose a scheduling algorithm that exhibits near optimal (on average within /spl sim/7% of optimum) delay/throughput performance in realistic network examples.

A compound model for TCP connection arrivals for LAN and WAN applications

We propose a two level model for TCP connection arrivals in local area networks. The first level are user sessions whose arrival is time-varying Poisson. The second level are connections within a user session. Their number and mean interarrival times are independent and biPareto across user sessions. The interarrivals within a user session are Weibull, and across all users are correlated Weibull. Our model has a small number of parameters which are inferred from real traffic collected at a firewall. We show that traffic synthesized with our model closely matches the original data. We extend this approach to a general model involving shot noise and show it is asymptotically consistent with more common fractal models used in data networks. Finally, we show that this model extends to the wide area network applications without alteration and it predicts smoothing of wide area network traffic profiles due to spatial aggregation, which we observe experimentally by synthetically creating a large aggregate TCP load.

Routing and protection in GMPLS networks: from shortest paths to optimized designs

Shortest path algorithms such as shortest path first (SPF) and constrained shortest path first (CSPF) are widely used in online traffic engineering where connections need to be set up one at a time as connection requests arrive sequentially. We propose an approach, called design-based routing (DBR), whereby optimized paths computed offline are used to guide online path setups. Offline path computation in generalized multiprotocol label switching (GMPLS) networks does not pose a significant challenge since optical core or metro networks typically consist of a few dozen to hundreds of nodes compared to hundreds to more than one thousand nodes in pure data networks. DBR takes advantage of available demand information based on customer prescriptions, traffic projections, and historical measurements to build an approximate traffic demand matrix for path optimization. By means of simulation, we perform comparative evaluations of opaque GMPLS networks under static and dynamic connections with different protection modes. The results indicate that DBR outperforms SPF and CSPF under a wide range of operating conditions and is robust to inaccuracies in the estimation of the traffic demand matrix. We then construct routing schemes with resource management and online measurement. The simulation results indicate that resource management provides an effective way to mitigate greed inherent in CSPF, and online measurement provides an effective way to improve DBR performance when the traffic demand information used in the design of DBR paths is different from the actual traffic demand.

Load characterization and anomaly detection for voice over IP traffic

We consider the problem of traffic anomaly detection in IP networks. Traffic anomalies typically arise when there is focused overload or when a network element fails and it is desired to infer these purely from the measured traffic. We derive new general formulae for the variance of the cumulative traffic over a fixed time interval and show how the derived analytical expression simplifies for the case of voice over IP traffic, the focus of this paper. To detect load anomalies, we show it is sufficient to consider cumulative traffic over relatively long intervals such as 5 min. We also propose simple anomaly detection tests including detection of over/underload. This approach substantially extends the current practice in IP network management where only the first-order statistics and fixed thresholds are used to identify abnormal behavior. We conclude with the application of the scheme to field data from an operational network.

SPIDER: A simple and flexible tool for design and provisioning of protected lightpaths in optical networks

Optical devices are poised to form the core of the next generation of backbone and enterprise networks. Optical routers, dense wavelength division multiplexing (DWDM) systems, and cross connects of unprecedented capacities are on the verge of large-scale commercial deployment. This massive buildup of optical gear necessitates careful planning and provisioning of the basic units of transmission — the light-paths. This paper presents a range of techniques for efficient and reliable optical network design, covering decentralized dedicated protection to shared path-based mesh restoration. These algorithms have been incorporated into SPIDER, an extensible software tool with a browser-based user interface, Java∗-based visualization, and spreadsheet input/output capabilities. We describe SPIDER and report on recent core networking applications using this tool, which also illustrate the key tradeoffs in optical network designs involving a variety of grades of protection and the balance between efficient use of wavelengths and restoration time.

Dynamic optimization in future cellular networks

With multiple air-interface support capabilities and higher cell densities, future cellular networks will offer a diverse spectrum of user services. The resulting dynamics in traffic load and resource demand will challenge present control loop algorithms. In addition, frequent upgrades in the network infrastructure will substantially increase the network operation costs if done using current optimization methodology. This motivates the development of dynamic control algorithms that can automatically adjust the network to changes in both traffic and network conditions and autonomously adapt when new cells are added to the system. Bell Labs is pursuing efforts to realize such algorithms with research on near-term approaches that benefit present third-generation (3G) systems and the development of control features for future networks that perform dynamic parameter adjustment across protocol layers. In this paper, we describe the development of conceptual approaches, algorithms, modeling, simulation, and real-time measurements that provide the foundation for future dynamic network optimization techniques.