Soumendra Nanda

Congestion-aware caching and search in information-centric networks

The performance of in-network caching in information-centric networks, and of cache networks more generally, is typically characterized by network-centric performance metrics such as hit rate and hop count, with approaches to locating and caching content evaluated and optimized for these metrics. We believe that user-centric performance metrics, in particular the delay from when a content request is made by the user to the time at which the requested content has been completely downloaded, are also important. For such metrics, performance is often determined by link capacity constraints and network congestion. We investigate network cache management and search policies that account for path-level (content-server to content-requestor) congestion and file popularity in order to directly minimize user-centric, content-download delay. Through simulation, we find that our policies yield significantly better download delay performance than existing policies, even though these existing policies provide better performance according to traditional metrics such as cache hit rate and hop count.

Localized Bridging Centrality for Distributed Network Analysis

Centrality is a concept often used in social network analysis to study different properties of networks that are modeled as graphs. We present a new centrality metric called localized bridging centrality (LBC). LBC is based on the bridging centrality (BC) metric that Hwang et al. recently introduced. Bridging nodes are nodes that are strategically located in between highly connected regions. LBC is capable of identifying bridging nodes with an accuracy comparable to that of the BC metric for most networks. As the name suggests, we use only local information from surrounding nodes to compute the LBC metric, whereas, global knowledge is required to calculate the BC metric. The main difference between LBC and BC is that LBC uses the egocentric definition of betweenness centrality to identify bridging nodes, while BC uses the sociocentric definition of betweenness centrality. Thus, our LBC metric is suitable for distributed or parallel computation and has the benefit of being an order of magnitude faster to calculate in computational complexity. We compare the results produced by BC and LBC in three examples. We applied our LBC metric for network analysis of a real wireless mesh network. Our results indicate that the LBC metric is as powerful as the BC metric at identifying bridging nodes. The LBC metric is thus an important tool that can help network administrators identify critical nodes that are important for the robustness of the network in a distributed manner.

Mesh-Mon: A multi-radio mesh monitoring and management system

Mesh networks are a potential solution for providing communication infrastructure in an emergency. They can be rapidly deployed by first responders in the wake of a major disaster to augment an existing wireless or wired network. We imagine a mesh node with multiple radios embedded in each emergency vehicle arriving at the site to form the backbone of a mobile wireless mesh. The ability of such a mesh network to monitor itself, diagnose faults and anticipate problems are essential features for its sustainable operation. Typical SNMP-based centralized solutions introduce a single point of failure and are unsuitable for managing such a network. Mesh-Mon is a decentralized monitoring and management system designed for such a mobile, rapidly deployed, unplanned mesh network and works independently of the underlying mesh routing protocol. Mesh-Mon nodes are designed to actively cooperate and use localized algorithms to predict, detect, diagnose and resolve network problems in a scalable manner. Mesh-Mon is independent of the underlying routing protocol and can operate even if the mesh routing protocol completely fails. One novel aspect of our approach is that we employ mobile users of the mesh, running software called Mesh-Mon-Ami, to ferry management packets between physically-disconnected partitions in a delay-tolerant-network manner. The main contributions of this paper are the design, implementation and evaluation of a comprehensive monitoring and management architecture that helps a network administrator proactively identify, diagnose and resolve a range of issues that can occur in a dynamic mesh network. In experiments on Dart-Mesh, our 16-node indoor mesh testbed, we found Mesh-Mon to be effective in quickly diagnosing and resolving a variety of problems with high accuracy, without adding significant management overhead.

Experiences with network coding within MANET field experiments

While network coding offers many theoretical properties that can be used to improve performance, practical experiences have been largely limited to simulation, emulation, and small-scale demonstrations. This paper reviews observations and results from recently conducted 802.11-based MANET field trials contrasting a network-coded protocol stack against a candidate baseline protocol suite. These field trials were based on an operational scenario with dismounts, vehicles, and airborne relay nodes conducting a search and rescue mission using video, chat, and situational awareness applications. This paper examines practical implementation issues such as CPU performance demands of network coding on an embedded system as well as observed traffic characteristics.

Multipath location aided routing in 2d and 3d

There are several ad hoc routing algorithms that utilize position information in two dimensional (2D) coordinates to make routing decisions at each node. The well known 2D position based routing algorithm, greedy perimeter stateless routing (GPSR) does not function correctly in three dimensions (3D). We designed MLAR (multipath location aided routing), a multipath variant of LAR (location aided routing). MLAR is position based, works efficiently in both 2D and 3D and uses alternate path routing. We have implemented and validated MLAR through simulations using ns-2 and studied its efficiency, scalability and other properties. We evaluated and compared the relative performance and benefits of our MLAR approach versus LAR, AODV and AOMDV in both 2D and 3D for a range of traffic and mobility scenarios. We present a direct comparison of a multipath position based algorithm with a multipath non-position based algorithm, as well as with both their single path versions. We also extended ns-2 to support 3D mobility models and routing protocols

Social network analysis plugin (SNAP) for mesh networks

In a network, bridging nodes are those nodes that from a topological perspective, are strategically located between highly connected regions of nodes. Thus, they have high values of the Bridging Centrality (BC) metric. We recently introduced the Localized Bridging Centrality (LBC) metric, which can identify such nodes via distributed computation, yet has an accuracy equal to that of the centralized BC metric. The LBC and BC metrics are based on the Social Network Analysis (SNA) metric “betweenness centrality”. We now introduce a new SNA metric that is more suitable for use in wireless mesh networks: the Localized Load-aware Bridging Centrality (LLBC) metric. The LLBC metric improves upon LBC by detecting critical bridging nodes while taking into account the actual traffic flows present in a mesh network. We only use local information from surrounding nodes to compute the LLBC metric, thus our LLBC metric is designed for scalable distributed computation and distributed network analysis. We developed the SNA Plugin (SNAP) for the Optimized Link State Routing (OLSR) protocol to study the potential use of LBC and LLBC in improving multicast communications. We present some promising initial results for SNAP from real and emulated mesh networks. SNAP is open source and free for academic use.

Capacity of Cache Enabled Content Distribution Wireless Ad Hoc Networks

While wireless ad hoc networks have a wide range of applications in environment monitoring, military operations, and disaster recovery, etc, the full potential of such networks is inherently hindered by their diminishing capacity as the network size scales up. Content caching has been previously proposed to improve the availability of contents in a network and thus helps to alleviate the load on content custodians, reduce access latency, and improve the network capacity. This paper studies the scaling laws of the capacity of cache-enabled content distribution wireless ad hoc networks. We consider two basic content access schemes, namely, the Nearest Caching Node scheme where a request is satisfied by the nearest node to the requestor that has the content in its cache, and the Transparent Enroute Caching scheme where a content request is routed towards the content custodian and is satisfied by an intermediate node (or custodian) along the path that has the content in its cache. We first establish the capacity of content distribution wireless ad hoc networks without content caching as a baseline for investigating the benefit of caching. We then obtain the scaling laws of the capacity for the above two content access schemes with content caching. Based on the results we further explore their design and performance implications. Our results show that the capacity exhibits distinct scaling behaviors under different scaling regimes of the network parameters. Under certain conditions increasing the cache size of the nodes can effectively improve the capacity while under other conditions the improvement can be negligible. The characterizations of the capacity allows us to identify the bottleneck of the content access capacity for given network scenarios and choose effective approaches to improve the capacity.

CONCERTO: Experiences with a Real-World MANET System Based on Network Coding

Publisher Summary This chapter presents control over network coding for enhanced radio transport optimization (CONCERTO), a fully implemented communication system based on a network coding protocol stack and reviews observations and results from recently conducted 802.11-based mobile ad hoc networks (MANET) field trials contrasting the CONCERTO system against a candidate baseline protocol suite. These field trials were based on operational scenarios with mobile radios conducting a search and rescue mission using video, file transfer, chat, and situational awareness applications. The CONCERTO system was shown to support 2 to 3 times more video throughput than a state-of-the-art set of protocols, as well as up to 7 times distance-utility product. This chapter examines practical implementation issues of network coding on an embedded system and analyzes performance results. CONCERTO derives its performance benefits from the combination of network coding and the use of a rich subgraph for forwarding. Network coding allows CONCERTO to exploit poor links and to use multiple forwarders to deliver information to mobile nodes. Network coded packets arriving from different forwarding nodes have a high probability of finding a path which allows delivery of sufficient information to recover the application data. Thus, CONCERTO achieves a high probability of delivery while efficiently using channel capacity. CONCERTO makes efficient use of the channel capacity by using multiple-paths to the destination and by using link-layer rather than end-to-end retransmissions to overcome link loss.

Cross-layer TCP adaptation in DISCO for tactical edge networks

In ongoing efforts, our team has developed DISCO (Distributed Cross-layer Architecture for Network Awareness and Opportunistic Transport), a cross-layer framework to enable network awareness and automated adaptation of transport, network and application parameters in a MANET. DISCO automates discovery of new applications and services, and applies rule-based logic and machine learning to enable the network to better meet applications performance requirements or for applications to adjust their requirements. DISCO also supports legacy applications using proxies to maintain backward compatibility. Our DISCO broker collects cross layer information, shares it in a distributed manner to guide new transport and network-layer adaptation methods. In this paper, we present an overview of DISCO and introduce a new TCP variant called TCP-Derwood-Lite that benefits from DISCOs design and provides performance comparable to the original TCP-Derwood. Unlike other TCP variants that operate on per packet microsecond timescales, Derwood-Lite operates on a per-second timescale in the user space, and still provides significant performance benefits in wireless networks with path or link switches. Our EMANE emulation experiments on engineering and tactical scenarios with as many as 23 mobile nodes in an OLSR MANET show 2X performance gains over TCP Reno and TCP Westwood, comparable to TCP Derwood but with much lower processing requirements.

Poster: Scalable evaluation for wireless mobile systems

Wireless researchers face the practical challenge of evaluating the performance of new routing protocols for very large-scale heterogeneous mobile systems. While simulation platforms, tools and techniques have matured, doubts remain about the broad validity of simulated results when applied to real world situations. Wireless network emulation provides a middle ground between simulations and real-world testing. However, emulation environments are complex to run and cost-prohibitive for larger-scale tests. In order to maximize fidelity while managing complexity and costs, we developed a common framework that unifies simulation-based analysis and emulation experiment tools, in order to produce calibrated and scalable performance evaluations.