Luoming Meng

Efficient Active Probing for Fault Diagnosis in Large Scale and Noisy Networks

Active probing is an effective tool for monitoring networks. By measuring probing responses, we can perform fault diagnosis actively and efficiently without instrumentation on managed entities. In order to reduce the traffic generated by probing messages and the measurement infrastructure costs, an optimal set of probes is desirable. However, the computational complexity for obtaining such an optimal set is very high. Existing works assume single-fault scenarios, apply only to small size networks, or use simplistic methods that are vulnerable to noises. In this paper, by exploiting the conditionally independent property in Bayesian networks, we prove a theorem on the information provided by a set of probes. Based on this theorem and structure property of Bayesian networks, we propose two approaches which can effectively reduce the computation time. A highly efficient adaptive probing algorithm is then presented. Compared with previous techniques, experiments have shown that our approach is more efficient in selecting an optimal set of probes without degraing diagnosis quality in large scale and noisy networks.

Probabilistic fault diagnosis for IT services in noisy and dynamic environments

The modern society has come to rely heavily on IT services. To improve the quality of IT services it is important to quickly and accurately detect and diagnose their faults which are usually detected as disruption of a set of dependent logical services affected by the failed IT resources. The task, depending on observed symptoms and knowledge about IT services, is always disturbed by noises and dynamic changing in the managed environments. We present a tool for analysis of IT services faults which, given a set of failed end-to-end services, discovers the underlying resources of faulty state. We demonstrate empirically that it applies in noisy and dynamic changing environments with bounded errors and high efficiency. We compare our algorithm with two prior approaches, Shrink and Maxcoverage, in two well-known types of network topologies. Experimental results show that our algorithm improves the overall performance.

QoE assessment and prediction method for high-definition video stream using image damage accumulation

The accuracy of the traditional assessment method of the quality of experience (QoE) has been facing challenges with the growth of high-definition (HD) video streaming services. Image display-quality damage is the main factor that affects the QoE in HD video services through UDP network transmission. In this paper, we introduce a novel objective factor known as image damage accumulation (IDA) to assess users QoE in HD video services. First, this paper quantitatively analyzed the effect on user quality of experience by IDA and established a mapping relationship between mean opinion scores and IDA. Furthermore, the probability of image damage caused by compression and transmission were analyzed. Based on this analysis, an objective QoE assessment and prediction method for HD video stream service that evaluated the user experience according to IDA are proposed. The proposed method can achieve assessment and prediction accuracy on three distinct subjective tests.

A max-flow/min-cut theory based multi-domain virtual network splitting mechanism

In network virtualization environment, if a virtual network (VN) needs to be deployed across multiple infrastructure domains, a splitting scheme of the VN should be found. With the goal of minimizing embedding cost, the existing methods solve VN splitting by linear programing. However, since the VN splitting problem is NP-Hard, these methods will take a lot of computing time when the problem scale gets bigger. In this paper, a max-flow/min-cut theory based VN splitting mechanism is proposed. The proposed method first creates a binary tree of the InPs by system clustering method, based on which the multidomain VN splitting problem is decomposed into several two-domain VN splitting problems. Then the method transforms each two-domain splitting problem into a max-flow/min-cut problem, and solves it by the shortest augmenting path algorithm efficiently. Simulations show that the proposed mechanism can improve the efficiency of VN splitting steadily and save the embedding cost.

A random switching traffic scheduling algorithm in wireless smart grid communication network

One of the key technologies of smart grid is an efficient, reliable and secure two-way communication system for meter data collection. Because of the advantages of muti-hop communication, self-organizing, self-healing and reliability, wireless muti-hop communication technology becomes an ideal choice for smart grid meter data collection. However, forming wireless mesh network with advanced electricity devices (smart meters) which have the communication capabilities for meter data collection faces challenge on communication performance of network caused by application layer data traffic. When a large number of data occur in emergence, some smart meters (the last hop nodes) which are in pivotal location will face great communication pressure and probably lead to extremely data congestion. With the idea of load balancing, this paper proposes a new random switching traffic scheduling algorithm based on meter data collection tree. Simulation data show that the new algorithm can create a balanced meter data collection tree, significantly reduce the packet loss ratio of the burst data and release congestion of system.

An improved network performance anomaly detection and localization algorithm

In this paper, we introduce a network performance anomaly detection and localization method based on active probing, aiming at avoiding waste of unnecessary probes and reducing detecting time by decreasing selecting rounds in detection phase. We propose a method of classifying detection strategies in order to find a balance between extra calculation and link load. Also we optimized the procedures of one of the strategies so that instead of finding a local optimal solution, we get a global optimal approach. An algorithm that can adapt to multi anomaly link networks is proposed and several issues during detection phase were being discussed. Finally we simulate a former representative algorithm and our improved method on different network topologies. The results show that our improved algorithm outperforms the former one in both probe selecting rounds during detection phase by 10%.

Preventing Congestion by Selective Admission Control in LTE-Based Public Safety Network

LTE-based Public Safety Network (PSN) is a wireless communication network which can provide efficient and reliable communication in disasters or emergencies for disaster relief and public protection. Therefore, ensuring that network congestion will not happen in PSN during an emergency is becoming increasingly important. LTE-based PSN is easy to be congested because part of spectrum resources is compressed to guarantee priority requirements of public safety users. In this paper, we develop a new method namely Selective Admission Control (SAC) mechanism to manage the radio bearers access to the commercial radio for Public Safety (PS) in LTE-based PSN. In the case of emergency, we select the traffic bearer with minimum estimated load increment accessing to the LTE-based PSN. The channel quality of new bearers should be taken into account, which means that in congestion, users who arrive earlier with poor channel quality will be rejected to reserve sufficient resources for users who arrive later with good channel quality. The simulation results show that the SAC mechanism can improve throughput by 36% and lower the rejection rate by 73% at most than reference method based on non-selective access control model, as a result effectively avoiding the network congestion and improving the utilization of spectrum resources for public safety communication.

User Association Based Cooperative Energy-Saving Mechanism in Heterogeneous 5G Access Networks

To reduce the energy consumption in heterogeneous 5G access networks, we propose a cooperative energy-saving mechanism based on user association (UA-CESM). Joint processing (JP) and centralized resource scheduling are introduced into the mechanism to solve inadequate coverage problem caused by base-station (BS) dormancy by means of cooperative BS selection and BS state control. First, the number of candidate dormant BSs is obtained on the basis of statistical traffic valley. Second, for heterogeneous cellular networks with relays, a BS selection model is designed to find cooperative BS set and a candidate dormant BS group. Finally, an energy-saving optimization strategy based on bipartite graph is proposed to realize user association from candidate dormant BSs to appropriate cooperative BSs. The simulation results show that the average rate of edge users is increased by 92.5%, and the coverage can be compensated without extra transmission power in the proposed mechanism.

RTagCare: Deep human activity recognition powered by passive computational RFID sensors

Activity recognition is a hot topic of research that is widely adopted by many applications such as fall detection of elderly people. Emerging passive RFID (radio-frequency identification) is creating huge opportunity for wearable devices to achieve activity recognition. However, performance of activity recognition is constrained by RFID localization accuracy and low quality of data streams characterized by sparsity and noise. In this paper, we present a novel activity recognition system, called RTagCare, which is a low-cost, unobtrusive and lightweight RFID based system. The RTagCare system leverage RFID localization technology, 3D-accelerometer base human activity identification and data mining algorithm to overcome traditional activity recognition system issues. RTagCare has been implemented and deployed in a test environment. As a result, RTagCare generally performs well to recognize human activity with high performance (F-score >94%).

Utility Maximization for Load Optimization in Cellular/WLAN Interworking Network Based on Generalized Benders Decomposition

Load steering is widely accepted as a key SON function in cellular/WLAN interworking network. To investigate load optimizing from a perspective of system utilization maximization more than just offloading to improve APs’ usage, a utility maximization (UTMAX) optimization model and an ASRAO algorithm based on generalized Benders Decomposition are proposed in this paper. UTMAX is to maximize the sum of logarithmic utility functions of user data rate by jointly optimizing user association and resource allocation. To maintain the flexibility of resource allocation, a parameter is added to the utility function, where smaller means more resources can be allocated to edge users. As a result, it reflects a tradeoff between improvements in user throughput fairness and system total throughput. UTMAX turns out to be a mixed integer nonlinear programming, which is intractable intuitively. So ASRAO is proposed to solve it optimally and effectively, and an optional phase for expediting ASRAO is proposed by using relaxation and approximation techniques, which reduces nearly 10% iterations and time needed by normal ASRAO from simulation results. The results also show UTMAX’s good effects on improving WLAN usage and edge user throughput.