Switched networks use meshed topologies for path redundancy; loop-free frame forwarding paths are established by protocols that use either the spanning or Dijkstra tree algorithms. When a network link or device fails, the tree must be repaired, leading to high convergence latencies impacting both broadcast and unicast traffic. Unicast frames are constrained to use the tree paths, which may not be the shortest path and links are not utilized uniformly. We propose a new meshed tree algorithm and meshed tree protocol (MTP) to provide non-looping paths (with backup) for broadcast frames. We also propose a new scheme for unicast frames to use shortest paths (provisioned with backup) between end devices. A novel virtual ID technique to construct the trees makes the protocol robust and provides fast failure recovery. In this article, we present implementation details of MTP on the GENI testbed and compare its performance with state-of-the-art loop-avoidance protocols.

Switched networks require loop-avoidance protocols to avoid looping of broadcast and multicast Loop-avoidance protocols use a tree algorithm to construct a logical tree from one root such as rapid spanning tree protocol (RSTP). RSTP results in high convergence latency and heavily impaired network performance on root switch failures. Some protocols construct a logical tree from every switch resulting in high operational overhead. We propose the Meshed Tree Protocol (MTP) to designate the number of roots based on desired network uptime and computational overhead, where a meshed tree conceptually maintains multiple tree branches from a root. MTP based on multi meshed tree algorithm (MMTA), constructs and maintains multiple meshed trees from multiple designated roots. We compare a prototype of MTP based on MMTA against RSTP using switched networks in the Global Environment for Network Innovation (GENI) testbed. MTP's root failure recovery response is several magnitudes superior to RSTP.

The Normalized Cut (NCut) objective function, widely used in data clustering and image segmentation, quantifies the cost of graph partitioning in a way that biases clusters or segments that are balanced towards having lower values than unbalanced partitionings. However, this bias is so strong that it avoids any singleton partitions, even when vertices are very weakly connected to the rest of the graph. Motivated by the Bühler-Hein family of balanced cut costs, we propose the family of Compassionately Conservative Balanced (CCB) Cut costs, which are indexed by a parameter that can be used to strike a compromise between the desire to avoid too many singleton partitions and the notion that all partitions should be balanced. We show that CCB-Cut minimization can be relaxed into an orthogonally constrained lt-minimization problem that coincides with the problem of computing Piecewise Flat Embeddings (PFE) for one particular index value, and we present an algorithm for solving the relaxed problem by iteratively minimizing a sequence of reweighted Rayleigh quotients (IRRQ). Using images from the BSDS500 database, we show that image segmentation based on CCB-Cut minimization provides better accuracy with respect to ground truth and greater variability in region size than NCut-based image segmentation.

The work in this paper studies the performance of Carrier Sense Multiple Access/Collision Avoidance (CSMA/CA) medium access control (MAC) mechanism in IEEE 802.11 DCF. From studying the propagation delay and clock synchronization differences between terminals, it is concluded that the slight timing differences, effectively a relative jitter, contributes to reducing the probability of collisions after a MAC backoff. This lesson is leveraged in a presented new CSMA/CA MAC technique that deliberately introduces a random short delay, akin to jitter, in the backoff slots time structure. The new medium access technique is evaluated through both an analytical model and simulations that consider the realistic timing constraints set in the IEEE 802.11 standard. Simulation results show improvements in normalized saturated throughput with the new technique from at least 14% for 10 nodes up to 26% for 50 nodes.

The mobile user population connected to the Internet continues to grow in number and hence provisioning seamless Internet service to such user is a topic of major interest. With new mobility architectures being investigated, it is important to have consistent comparative platforms and metrics. In this article, we introduce a novel and user-centric handoff cost framework to analyze handoff performance of different mobility schemes. The proposed framework helps examine the impacts of registration costs, signaling overhead, and data loss for Internet connected mobile users employing a unified cost metric. We first applied the framework to our future Internet mobility architecture, called Virtual Mobility Domains (VMDs). This architecture is user centric as it allows mobile users to avail services from overlapping mobility domains based on their mobility profiles. Next, the framework is applied to IPv6-based mobility protocols such as Hierarchical Mobile IPv6 and Proxy Mobile IPv6 to show the framework's flexibility and adaptability. Using the framework, we compare the handoff performance of IPv6-based mobility protocols to the VMD based mobility scheme. The results indicate that the handoff performance achieved with VMD is three and nine times superior to Proxy Mobile IPv6 and Hierarchical Mobile IPv6 respectively.