ODMCA: An adaptive data mining control algorithm in multicarrier networks

Multicarrier communication is a promising technique to effectively deliver high data rate and combat delay spread over fading channel, and adaptability is an inherent advantage of multicarrier communication systems. It can be implemented in online data streams. This paper addresses a significant problem in multicarrier networks that arises in data streaming scenarios, namely, today’s data mining is ill-equipped to handle data streams effectively, and pays little attention to the network stability and the fast response [http://www-db.standford.edu/stream]. Furthermore, in analysis of massive data streams, the ability to process the data in a single pass, while using little memory, is crucial. For often the data can be transmitted faster than it can be stored or accessed from disks.To address the question, we present an adaptive control-theoretic explicit rate (ER) online data mining control algorithm (ODMCA) to regulate the sending rate of mined data, which accounts for the main memory occupancies of terminal nodes. This single-pass scheme considers limited memory space to process dynamic data streams, and also explores the adaptive capability, which is employed in a general network computation model for dynamic data streams. The proposed method uses a distributed proportional integrative plus derivative (PID) controller combined with data mining, where the control parameters can be designed to ensure the stability of the control loop in terms of sending rate of mined data. The basic PID approach for the computation network transmission is presented and z-transformation and Schur–Cohn stability test are used to achieve the stability criterion, which ensures the bounded rate allocation without steady state oscillation. We further show how the ODMCA can be used to design a controller, analyze the theoretical aspects of the proposed algorithm and verify its agreement with the simulations in the LAN case and the WAN case. Simulation results show the efficiency of our scheme in terms of high main memory occupancy, fast response of the main memory occupancy and of the controlled sending rates.     

Orthogonal margin discriminant projection for dimensionality reduction

Dimensionality reduction aims to represent high-dimensional data with much smaller number of features, which plays as a preprocessing step to remove the insignificant and irrelevant features in many machine learning applications, resulting in lower computational cost and better performance of classifiers. In most cases, the data points can be well classified with margin samples which are defined as furthest intra-class samples and nearest inter-class samples. Motivated by this observation, this paper proposes a linear supervised dimensionality reduction method called orthogonal margin discriminant projection (OMDP). After OMDP projection, intra-class data points become more compact and inter-class data points become more separated. Extensive experiments have been conducted to evaluate the proposed OMDP algorithm using several benchmark face data sets. The experimental results confirm the effectiveness of the proposed method.     

A Fast Formation Flocking Scheme for a Group of Interactive Distributed Mobile Nodes in Autonomous Networks

Ubi-Com promises to provide a diversity of services at anytime and anyplace. The mobile nodes are envisioned to cooperate freely and move on a plane with the help of the efficient control and distributed coordination strategies. To achieve this goal, mobile nodes group communications in Ubi-Com is becoming increasing important and challenging. In this paper, we focus on the formation flocking in a group of autonomous mobile nodes. In the scenario, there are two kinds of nodes: the leader node and the follower nodes. The follower nodes are required to follow the leader node wherever it goes (following), while keeping a particular formation they are given in input (flocking). A novel scheme is proposed on the basis of the relative motion theory. Extensive theoretical analysis and simulation results have demonstrated that this scheme provides the follower nodes an efficient and timely method to follow the leader with the shortest path and the shortest time. In addition, the reported scheme is scalable in the sense that the processing load in each node is not increasing with more nodes in a group.     

TPGF: geographic routing in wireless multimedia sensor networks

In this paper, we propose an efficient Two-Phase geographic Greedy Forwarding (TPGF) routing algorithm for WMSNs. TPGF takes into account both the requirements of real time multimedia transmission and the realistic characteristics of WMSNs. It finds one shortest (near-shortest) path per execution and can be executed repeatedly to find more on-demand shortest (near-shortest) node-disjoint routing paths. TPGF supports three features: (1) hole-bypassing, (2) the shortest path transmission, and (3) multipath transmission, at the same time. TPGF is a pure geographic greedy forwarding routing algorithm, which does not include the face routing, e.g., right/left hand rules, and does not use planarization algorithms, e.g., GG or RNG. This point allows more links to be available for TPGF to explore more routing paths, and enables TPGF to be different from many existing geographic routing algorithms. Both theoretical analysis and simulation comparison in this paper indicate that TPGF is highly suitable for multimedia transmission in WMSNs.