Guo-Kai Ni

Resource Allocation with Interference Avoidance in OFDMA Femtocell Networks

Interference control and quality-of-service (QoS) awareness are the major challenges for resource management in orthogonal frequency-division multiple access femtocell networks. This paper investigates a self-organization strategy for physical resource block (PRB) allocation with QoS constraints to avoid the co-channel and co-tiered interference. Femtocell self-organization including self-configuration and self-optimization is proposed to manage the large femtocell networks. We formulate the optimization problem for PRB assignments where multiple QoS classes for different services can be supported, and interference between femtocells can be completely avoided. The proposed formulation pursues the maximization of PRB efficiency. A greedy algorithm is developed to solve the resource allocation formulation. In the simulations, the proposed approach is observed to increase the system throughput by over 13% without femtocell interference. Simulations also demonstrate that the rejection ratios of all QoS classes are low and mostly below 10%. Moreover, the proposed approach improves the PRB efficiency by over 82% in low-loading scenario and 13% in high-loading scenario.

Top-

Storage nodes are expected to be placed as an intermediate tier of large scale sensor networks for caching the collected sensor readings and responding to queries with benefits of power and storage saving for ordinary sensors. Nevertheless, an important issue is that the compromised storage node may not only cause the privacy problem, but also return fake/incomplete query results. We propose a simple yet effective dummy reading-based anonymization framework, under which the query result integrity can be guaranteed by our proposed verifiable top- k query (VQ) schemes. Compared with existing works, the VQ schemes have a fundamentally different design philosophy and achieve the lower communication complexity at the cost of slight detection capability degradation. Analytical studies, numerical simulations, and prototype implementations are conducted to demonstrate the practicality of our proposed methods.

k

The complexity that telecommunications companies are faced with in their business processes and their information technology (IT) systems is especially apparent in their billing systems. These systems are required not only to handle large volumes of data and frequent changes in business rules, but also to ensure that the billing be done accurately and on time. This paper describes a solution that was developed to address this problem. It consists of an operations support system that is compliant with NGOSS (Next Generation Operations System and Software) and it implements a service-oriented architecture (SOA) that relies on an enhanced enterprise service bus (ESB). This enhanced ESB, referred to here as an adaptable service bus (ASB), makes it possible to carry out changes to business rules at runtime, thus avoiding costly shutdowns to the billing application. An implementation of this system has been operational in ChungHwa Telecom Company, Taiwan, since January 2008 and provides complete support to its billing application. As a result, the billing process cycle time has been reduced from 10–16 days to 3–4 days, which cleared the way for further growth of the business.

Query Result Completeness Verification in Tiered Sensor Networks

In large scale sensor networks, storage nodes may be placed as an intermediate tier for caching the sensor readings and respond to queries, resulting in reduced power consumption and storage savings for sensors. However if a storage node is compromised, it will create privacy issues and may return fake/incomplete results. Thus we suggest a simple yet effective dummy reading-based anonymization framework, whereby query result integrity can be guaranteed by our proposed Verifiable top-k Query (VQ) scheme. Compared to existing work, VQ schemes have a different design philosophy and achieve the lowest communication complexity, with only constant overhead, at with a small degradation of detection capability. Analysis and numerical examples demonstrate the practicality of our proposed approach.

A runtime-adaptable service bus design for telecom operations support systems

The ad hoc network is composed of multiple sensor nodes to serve various applications, such as data collection or environmental monitoring. In many applications, the sensor nodes near the boundary of the deployment region provide biased or low-quality information because they have limited number of neighboring nodes and only partial information is available. Hence, the boundary recognition is an important issue in the ad hoc networks. By the statistical approach in high node density networks, Feketes pioneer work identified the boundary node by number of neighboring nodes and using a specific threshold. By exploiting the number of nodes in the two-hop region, our proposed algorithm has significant improvement of boundary recognition contrasted with Feketes algorithm in the low-density network. Given the information topology and the cost function, the analyses provide a framework to obtain the optimal threshold for boundary recognition. Besides, the simulation results reveal the proposed algorithm has greater than 90% detection rate and lower than 10% false alarm rate.

Top-k query result completeness verification in sensor networks

A new efficient algorithm using the compressed domain features of H.264 INTRA frames is proposed for moving object extraction on huge video surveillance archives. To achieve searching efficiency, we propose to locate moving objects by scrutinizing only the INTRA frames in video surveillance archives in H.264 compressed domain with short GOP length. In the proposed structure, a modified codebook algorithm is designed to build the block-based background models from the INTRA coding features. Through the subtraction with the background codebook models, the foreground energy frame is filtered and normalized for detecting the existence of moving objects. To overcome the over-segmentation problem and enable the unsupervised searching, a new structure of hysteresis thresholding, where the thresholds are obtained automatically by an efficient algorithm, is adopted to extract foreground blocks. At the final step, the connected components labeling (CCL) and morphological filters are employed to obtain the list of moving objects. As shown in the experimental results, the proposed algorithm outperforms representative existing works.

Optimal self boundary recognition with two-hop information for ad hoc networks

Information systems are among the tools most frequently used by businesses in delivering and maintaining services. Yet the challenges faced by enterprises using information technology are not exclusively linked to system development, but are also derived from system operation and management. This paper describes a solution that was developed to address the importance of service operation management. The framework presented here was implemented by the Chunghwa Telecom Company in an effort to improve information system effectiveness and to reduce costs of system operation management. Since July 2008, the system has been providing complete support of the daily operations of the companys billing system. In consequence of the systems implementation, the company has experienced a large-scale improvement in both efficiency and cost reduction. Additionally, the number of incident occurrences has been reduced from an average of 50 monthly to 15. The average amount of time spent addressing individual incidents has also been reduced from approximately 20 hours to 26 minutes.

Moving Object Extraction Using Compressed Domain Features of H.264 INTRA Frames

The Domain Name System (DNS) is one of the critical components of modern Internet networking. Proper Internet functions (such as mail delivery, web browsing and so on) are typically not possible without the use of DNS. However with the growth and commercialization of global networking, this protocol is often abused for malicious purposes which negatively impacts the security of Internet users. In this paper we perform security data analysis of DNS traffic at large scale for a prolonged period of time. In order to do this, we developed DNSPacketlizer, a DNS traffic analysis tool and deployed it at a mid-scale Internet Service Provider (ISP) for a period of six months. The findings presented in this paper demonstrate persistent abuse of the protocol by Botnet herders and antivirus software vendors for covert communication. Other suspicious or potentially malicious activities in DNS traffic are also discussed.

A service-oriented management framework for telecom operation support systems

The complexity that telecommunications companies are faced with in their Operations Support Systems (OSSs) is especially apparent in their billing systems. These systems are required to handle large volumes of data located on different legacy systems operating on heterogeneous platforms. These systems are often integrated using service oriented architecture (SOA). The successful completion of billing processes is essential to the operation of a telecommunications company, as it ensure the smooth inflow of income from services rendered. When a billing process fails it is the generally the result of one of two types of faults - availability faults or reliability faults. Unless the system has been designed with fault-tolerance in mind, failures constitute an extremely costly and time consuming occurrence. This paper proposes a Service Oriented Fault Tolerant Environment (SOFTEN) that seeks to address this problem directly. The SOFTEN described here is composed of three key modules: a Fault Tolerant Proxy (FT Proxy), Fault Tolerant Services (FT Services), and an Adaptable Service Bus (ASB). The FT Proxy allows developers to define strategies and responses for dealing with the occurrence of both availability and reliability faults. FT Services are fault handling mechanisms that enact the procedures established in the FT Proxy. The ASB provides intelligent routing and data transform between the billing process and the billing services. Further, the paper presents results that demonstrate the effectiveness of the solution, both in increasing reliability and in reducing calculation time for completion of the billing process. An implementation of this system has been operational at the ChungHwa Telecom Company, in Taiwan since May 2008 and provides complete support for its billing application. As a result, the billing process cycle time has been reduced from 10-16 days to 3-4 days, which cleared the way for further growth of the business.

Investigating DNS traffic anomalies for malicious activities

For electromagnet interference (EMI) optimization issues, different hardware-level techniques have been proposed. This paper focuses on an EMI optimization technique via a software-level technique. We propose a novel estimation and optimization tool for reducing conducted EMI at specified frequency by compiler technology. This study is not a research on compiler technique but an adaption of computer science-domain technology to EMI optimization research. The proposed tool can accept C language syntax and generate many versions of assembly programs. These assembly programs perform the same functionality defined by the input C program but with different conducted EMI behaviors when they are executed. The proposed tool can estimate, select, and generate the assembly program with the least amount of conducted EMI released during its execution. The experiment results show that the proposed tool can analyze test C-programs and generate lower EMI assembly programs. Currently, compared to a commercial compiler, the proposed technique can decrease conducted EMI by 2-5 dB at any specified frequency.