Marcus Pang

Automatic Identification of Impairments Using Support Vector Machine Pattern Classification on Eye Diagrams

We have demonstrated powerful new techniques for identifying the optical impairments causing the degradation of an optical channel. We use machine learning and pattern classification techniques on eye diagrams to identify the optical impairments. These capabilities can enable the development of low-cost optical performance monitors having significant diagnostic capabilities

Scalable request routing with next-neighbor load sharing in multi-server environments

Load balancing for distributed servers is a common issue in many applications and has been extensively studied. Several distributed load balancing schemes have been proposed that proactively route individual requests to appropriate servers to best balance the load and shorten request response time. These schemes do not require a centralized load balancer. Instead, each server is responsible for determining, for each request it receives from a client, to which server in the pool the request should be forwarded for processing. We propose a new request routing scheme that is more scalable to increasing number of servers and request load than the existing schemes. The method combines random server selection and next-neighbor load sharing techniques that together prevent the staleness of load information from building up when the number of servers increases. Our simulation shows that it outperforms existing schemes under a piggyback-based load update model.

An integrated soft handoff approach to IP fast reroute in wireless mobile networks

This paper presents an integrated approach to IP fast reroute (IPFRR) of both unicast and multicast paths in wireless mobile networks. A distinct feature of the proposed approach is that, instead of modifying existing routing protocols, it employs a soft handoff technique, i.e., temporarily installs pre-computed Loop Free Alternative Paths (LFAPs) until the co-existing routing protocol converges to new routes. The proposed approach builds on our previously proposed IPFRR technology and uses the concept of pre-computed LFAPs not only for local but also for remote link failures within a certain neighborhood to achieve full alternative path coverage for a single link failure. This papers contributions include: i) bandwidth efficient fast failure detection by integrating two novel mechanisms, namely probing and link quality prediction, ii) a novel method for calculating LFAPs, iii) a framework for switching seamlessly between LFAPs and OSPF paths, iv) a multicast fast reroute mechanism, and v) implementation in eXtensible Open Router Platform (XORP). We also present a generic framework for handling multiple simultaneous failures in the integrated IPFRR. The performance evaluation has been performed in both indoor and outdoor environments with real 802.11 radio links. The results confirm that our IPFRR technology consistently provides significant convergence time improvement during a single link failure event.

Context-driven disruption tolerant networking for vehicular applications

In this paper, we describe how a context-driven disruption tolerant network may be utilized for vehicular applications. The key concept involves utilizing contextual information pertaining to space and time settings of applications to drive networking functions. As a result, the performance characteristics may be better attuned to the time-scale of relevant vehicular applications. We discuss the protocol highlights and present a compact grid-based method for performance evaluation of such systems. Novel metrics such as information freshness demonstrate the discrete nature of information dissemination. Additionally, we observe that asymptotic evaluation of metrics such as packet delivery ratio do not provide a realistic picture of application performance. As a consequence, the time-scale of interest for the applications becomes crucial.

Self-correcting Adaptive Tracking System

The benefits of Global Positioning System (GPS) are recognized in numerous military as well as civilian applications. In many situations, however, GPS signals are simply not available or, at best, intermittently observable. This paper describes a novel location tracking system, called self-correcting adaptive tracking system (SATS), which focuses on solving group location problem when GPS is not available. In our location tracking system, we use a tracking mechanism that allows locating group members based on their pair wise distance information. A key innovation of SATS is that we use an adaptive search algorithm to find the new position estimate based on constraints given by the ranged data. In addition, our location tracking system is capable of extracting directional information normally unavailable in ranging system, which allows us to adaptively stabilize the orientation of the group. The SATS methodology has been prototyped and tested as part of an Office of Navy Research (ONR) program.

Using an adaptive management plane for policy-based network management traffic in MANETs

Network management for Mobile Ad Hoc Networks (MANETs) is a challenging management problem given the intermittent connectivity of the nodes and the low bandwidth constraints associated with these networks. Further, MANET management mandates a distributed management paradigm, which gives rise to specific information dissemination challenges. In order to manage these networks, the Network Management System (NMS) needs to send critical network management alerts and data to network operation centers (NOCs) and the NOCs need to send changes to policies and configuration files to the distributed nodes. The need to keep the overhead of management traffic to a minimum and yet reliably deliver this data is a requirement for any NMS in this environment. This paper examines these challenges and proposes an Adaptive Management Plane approach that overcomes these challenges. This approach provides support for Disruption Tolerant Networking (DTN), allowing messages to reach intermittently connected nodes. It also provides a service to deliver management data to the remote nodes according to the information dissemination requirements that regulate the expiration, revision and confirmation of the data. In addition, the approach provides support for above Multi-Topology Routing (MTR), allowing the NMS to deliver data of different priorities over multiple networks that exhibit different traffic delivery characteristics. This solution is described in terms of an implementation in the Dynamic Re-Addressing and Management for the Army (DRAMA) policy-based network management system.

DiffServ QoS and OLSR MANET Outdoor Demonstration

The CECOM MOSAIC ATD encompasses an integrated set of diverse technologies to demonstrate rapidly deployable, secure, robust, assured-QoS communications among mobile ad hoc nodes. The assured IP QoS technology solution integrates DiffServ based QoS resource management with centralized admission control over a dynamic tactical network environment. We describe an outdoor demonstration of our QoS technology prototype over an on-the-move MANET running OLSR. This demonstration does not include IP mobility support, however provides an alternate route between the ad hoc nodes through the use of two distinct wireless networks. The demonstration exhibits route changes between single and multiple IP hops as the nodes drive, and shows handoff between wireless networks as they move out of range of one radio network to the other. We summarize our observations and empirical performance results for real-time traffic