Shing-Tsaan Huang

Detecting termination of distributed computations by external agents

An algorithm is presented that defects for termination of distributed computations by an auxiliary controlling agent. The algorithm assigns a weight W, 0<W<or=1, to each active process and to each message in transit. The algorithm maintains that the sum of all the weights related to the computation is equal to one. The controlling agent terminates the algorithm if its weight equals one. A space-efficient scheme is proposed to encode the weights such that an active process can send a very large number of messages before reaching a weight equal to one. Thus, in the proposed encoding scheme, each process and message needs only a small number of bits to encode the weight; the processes can be almost free from the delays of waiting for a supply of weights from the controlling agent.<<ETX>>

Adaptive splitting and pre-signaling for RFID tag anti-collision

In an RFID system a reader requests tags to send their IDs by RF signal backscattering for the purpose of identification. When multiple tags respond to the request simultaneously, tag collisions occur and the tag identification performance is degraded. There are several tag anti-collision protocols proposed for reducing tag collisions. The protocols can be categorized into two classes: ALOHA-based and tree-based protocols that include deterministic tree-based and probabilistic counter-based subclasses of protocols. ALOHA-based protocols have the tag starvation problem; deterministic tree-based protocols have the problem that their performances are influenced by the length and/or the distribution of tag IDs. On the contrary, probabilistic counter-based protocols do not have such problems. In this paper, we propose a probabilistic counter-based tag anti-collision protocol, called ASPS, to reduce tag collisions by adaptively splitting tags encountering collisions into several groups according to the estimated number of tags to be split, and to reduce the number of messages sent between the reader and tags by utilizing a pre-signaling bit. We simulate and analyze ASPS and compare it with related ones to show its advantages.

Cohorts structures for fault-tolerant k entries to a critical section

We propose a structure named Cohorts to solve the problem of the access control of multiple entries to a critical section. Our solution is formalized as forming quorums in a k-coterie. It is resilient to node failures and/or network partitioning, invokes constant expected message cost and has comparably high availability.

k-coteries for fault-tolerant k entries to a critical section

The authors extend the concept of coterie into k-coterie for k entries to a critical section. A structure named Cohorts is proposed to construct quorums in a k-coterie. The solution is resilient to node failures and/or network partitioning and has a low communication cost. The Cohorts structure is further improved to increase the availabilities of 1-entry critical sections.<<ETX>>

ALRD: AoA Localization with RSSI Differences of Directional Antennas for Wireless Sensor Networks

In this paper, we fit RSSI values into a parabola function of the AoA between 0° and 90° by applying quadratic regression analysis. We also set up two-directional antennas with perpendicular orientations at the same position and fit the difference of the signal RSSI values of the two antennas into a linear function of the AoA between 0° and 90° by linear regression analysis. Based on the RSSI-fitting functions, we propose a novel localization scheme, called AoA Localization with RSSI Differences (ALRD), for a sensor node to quickly estimate its location with the help of two beacon nodes, each of which consists of two perpendicularly orientated directional antennas. We apply ALRD to a WSN in a 10 × 10  m indoor area with two beacon nodes installed at two corners of the area. Our experiments demonstrate that the average localization error is 124 cm. We further propose two methods, named maximum-point minimum-diameter and maximum-point minimum-rectangle, to reduce localization errors by gathering more beacon signals within 1 s for finding the set of estimated locations of maximum density. Our results demonstrate that the two methods can reduce the average localization error by a factor of about 29%, to 89 cm.

Distributed execution model for self-stabilizing systems

There are several execution models for self-stabilizing systems discussed in the literature. Among them the distributed model is a more realistic one in the sense that it makes the weakest assumption about the execution environment; whereas the serial model is a less realistic one in the sense that it makes the strongest assumption. In this paper we first discuss how to convert a self-stabilizing system operating with the serial model into a system operating with the distributed model, but such a conversion does not guarantee that the converted system is self-stabilizing. Then we propose a transform technique which makes the proof whether or not the converted system is self-stabilizing much easier.<<ETX>>

A self-stabilizing (Δ+4)-edge-coloring algorithm for planar graphs in anonymous uniform systems

This paper proposes a self-stabilizing edge-coloring algorithm using (@D+4) colors for distributed systems of a planar graph topology, where @D>=5 is the maximum degree of the graph. The algorithm can be applied to anonymous uniform systems and its time complexity is O(n^2) moves under the central daemon model.

The Fuzzy Philosophers

Consider a network of nodes: each node represents a philosopher: links represent the neighboring relationship among tile philosophers, Every philosopher enjoys singing so much that once getting the chance, he always sings a song within a finite delay. Fhis paper proposes a protocol for the philosophers to follow. Fhe protocol guarantees the following requirements: (1) No two neighboring philosophers sing songs simultaneously. (2) Along any infinite time period, each philosopher gets his chances to sing infinitely often. Following tile protocol, each philosopher uses only one bit to memorize his state.Sometimes tile philosophers may be fuzzy enough to forget tile state. So, a self-stabili ing vesion of the protocol is also proposed to cope with this problem, However, the philosophers may need additional bits to memorize their states.

Obtaining nondominated k-coteries for fault-tolerant distributed k-mutual exclusion

A k-coterie is a family of sets (called quorums) in which any (k+1) quorums contain at least a pair of quorums intersecting each other. K-coteries can be used to develop distributed k-mutual exclusion algorithms that are resilient to node and/or communication link failures. A k-coterie is said to dominate another k-coterie if and only if every quorum in the latter is a super set of some quorum in the former. Obviously the dominating one has more chance than the dominated one for a quorum to be formed successfully in an error-prone environment. Thus, we should always concentrate on nondominated k-coteries that no k-coterie can dominate. We introduce a theorem for checking the nondomination of k-coteries, define a class of special nondominated k-coteries-strongly nondominated (SND) k-coteries, and propose two operations to generate new SND k-coteries from known SND k-coteries.

Fault-Tolerant and secure networked storage

This paper proposes a coding-based fault-tolerant and secure scheme for networked storage systems which is suitable for different architectures, such as peer-to-peer (P2P) networks, Grids and Clouds. Unlike replication-based schemes, the proposed scheme encodes and inserts into a data object with little redundancy, instead of using replicas of the whole object, to make the systems fault-tolerant. The scheme divides the data object into blocks, encodes the blocks, encrypts the blocks by light-weight XOR operations with a secret key, rearranges the encrypted blocks into pieces, and stores the pieces in separate computers or nodes. The scheme is fault-tolerant in the sense that the original object is available even if some of the pieces cannot be accessed due to node and/or link failures. The scheme is also secure in the sense that the data object can be reconstructed only if the correct secret key is provided. The scheme is compared with related ones to show its advantages.