It is extremely difficult to train a superhuman Artificial Intelligence (AI) for games of similar size to StarCraft II. AlphaStar is the first AI that beat human professionals in the full game of StarCraft II, using a league training framework that is inspired by a game-theoretic approach. In this paper, we improve AlphaStar's league training in two significant aspects. We train goal-conditioned exploiters, whose abilities of spotting weaknesses in the main agent and the entire league are greatly improved compared to the unconditioned exploiters in AlphaStar. In addition, we endow the agents in the league with the new ability of opponent modeling, which makes the agent more responsive to the opponent's real-time strategy. Based on these improvements, we train a better and superhuman AI with orders of magnitude less resources than AlphaStar (see Table 1 for a full comparison). Considering the iconic role of StarCraft II in game AI research, we believe our method and results on StarCraft II provide valuable design principles on how one would utilize the general league training framework for obtaining a least-exploitable strategy in various, large-scale, real-world games.

Federated learning (FL) is a well established method for performing machine learning tasks over massively distributed data. However in settings where data is distributed in a non-iid (not independent and identically distributed) fashion - as is typical in real world situations - the joint model produced by FL suffers in terms of test set accuracy and/or communication costs compared to training on iid data. We show that learning a single joint model is often not optimal in the presence of certain types of non-iid data. In this work we present a modification to FL by introducing a hierarchical clustering step (FL+HC) to separate clusters of clients by the similarity of their local updates to the global joint model. Once separated, the clusters are trained independently and in parallel on specialised models. We present a robust empirical analysis of the hyperparameters for FL+HC for several iid and non-iid settings. We show how FL+HC allows model training to converge in fewer communication rounds (significantly so under some non-iid settings) compared to FL without clustering. Additionally, FL+HC allows for a greater percentage of clients to reach a target accuracy compared to standard FL. Finally we make suggestions for good default hyperparameters to promote superior performing specialised models without modifying the the underlying federated learning communication protocol.

Anomaly detection is a problem with applications for a wide variety of domains; it involves the identification of novel or unexpected observations or sequences within the data being captured. The majority of current anomaly detection methods are highly specific to the individual use case, requiring expert knowledge of the method as well as the situation to which it is being applied. The Internet of Things (IoT) as a rapidly expanding field offers many opportunities for this type of data analysis to be implemented, however, due to the nature of the IoT, this may be difficult. This review provides a background on the challenges which may be encountered when applying anomaly detection techniques to IoT data, with examples of applications for the IoT anomaly detection taken from the literature. We discuss a range of approaches that have been developed across a variety of domains, not limited to IoT due to the relative novelty of this application. Finally, we summarize the current challenges being faced in the anomaly detection domain with a view to identifying potential research opportunities for the future.

Vehicular ad-hoc networks (VANETs) that enable communication among vehicles and between vehicles and un- manned aerial vehicles (UAVs) and cellular base stations have re- cently attracted significant interest from the research community, due to the wide range of practical applications they can facilitate (e.g. road safety, traffic management, pollution monitoring and rescue missions). Despite this increased research activity, the high vehicle mobility in a VANET raises concerns regarding the robustness and adaptiveness of such networks to support system applications. Instead of allowing direct communications between every vehicle to UAVs or base stations, clustering methods will potentially be efficient to overcome bandwidth, power consump- tion and other resource issues. Using the clustering technique, neighbouring vehicles are grouped into clusters with a particular vehicle elected as the Custer Head (CH) in each cluster. Each vehicle communicates with UAVs or base stations through the CH of the associated cluster. Despite the potential advantages, a major challenge for clustering techniques is to maintain cluster stability in light of vehicle mobility and radio fluctuation. In this paper, we propose a Stable Clustering Algorithm for vehicular ad hoc networks (SCalE). Two novel features are incorporated into the algorithm: knowledge of the vehicles behaviour for efficient selection of CHs, and the employment of a backup CH to maintain the stability of cluster structures. By simulation methods, these are shown to increase stability and improve performance when compared to existing clustering algorithms.

Recent research on Low Power Wide Area Network (LPWAN) technologies which provide the capability of serving massive low power devices simultaneously has been very attractive. The LoRaWAN standard is one of the most successful developments. Commercial pilots are seen in many countries around the world. However, the feasibility of large scale deployments, for example, for smart city applications need to be further investigated. This paper provides a comprehensive case study of LoRaWAN to show the feasibility, scalability, and reliability of LoRaWAN in realistic simulated scenarios, from both technical and economic perspectives. We develop a Matlab based LoRaWAN simulator to offer a software approach of performance evaluation. A practical LoRaWAN network covering Greater London area is implemented. Its performance is evaluated based on two typical city monitoring applications. We further present an economic analysis and develop business models for such networks, in order to provide a guideline for commercial network operators, IoT vendors, and city planners to investigate future deployments of LoRaWAN for smart city applications.

The enhancement of wireless technologies for machine-type communications and the Internet of things attracts the attention of recent communications research. The IEEE 802.11 task group TGah proposes to use group-based contention scheme in the MAC of its proposed amendment 11ah in order to cope with massive power constrained stations (STAs) per access point (AP). However the detailed regrouping technique is described as implementation specific. In this paper, we propose a novel method consisting of a signalling process and a regrouping algorithm. The aim of the regrouping is to minimise the potential transmission collisions caused by the hidden node problem. Firstly, the AP acquires knowledge of potential hidden node pairs and the traffic requirement of the STAs in the network. The AP then regroup the STAs into different contention groups according to either a centralised Viterbi-like algorithm or a decentralised iterative updating manner. The proposed technique can also be used in other contention based wireless networks where grouping is supported in the MAC protocol. System level simulation results show superior performance of the proposed technique in terms of the reduction in collision.

This paper proposes a dynamic user association method to address the load balancing problem in dense IEEE 802.11ax networks with uneven user distribution, where a user determines which AP to connect to taking into consideration of multiple factors such as RSS, potential relative capacity, achievable data rate and location of users. The method is user-centric and does not incur much signaling overhead, while performance optimization can be achieved without inter-AP coordination. Simulation results have been presented to show that the proposed solution can improve load balancing and have higher average throughput for 10% worst users as well as maintain the maximal total system throughput. Our evaluation also suggests that the location-awareness of users considered in the proposed solution plays an important role to improve the performance.

Activity recognition in smart environments is an important technology for assisted living and e-health. Recently there are growing interests in applying machine learning algorithms to activity recognition tasks. In this paper, we combine support vector machine (SVM) and association rule learning to improve the performance of activity recognition based on streaming sensor data in smart homes. The proposed approach allows us to accurately identify the activity boundaries, hence reducing activity recognition errors in the system.

The Non Line-of-Sight (NLoS) issue is considered as a key challenge to mature indoor localization technologies. Using Time of Arrival (TOA) on Ultra-wideband (UWB) signal is able to provide fine localization accuracy if NLoS-biased anchor nodes can be identified. In this paper, an antenna cluster selection scheme is proposed to identify the NLoS-biased node, therefore improve the localization accuracy and provide location-aware communication service. In particular, the area of interest with NLoS human body blockage is divided into multiple convex hull based effective localization areas (ELAs). Four antennas, from a distributed antenna system, are deployed in each ELA, where the most severely NLoS-biased antenna is identified and discarded, two other antennas are selected to execute a 2-dimension (2D) indoor localization. The remaining antenna uses this location information to provide location-aware communication service. Performance is evaluated using a channel sounder's measurement data and a simulator, benchmarked to the traditional parametric identification method.

Pseudo medium/low voltage (MV/LV) transformer loads are usually used as partial inputs to the distribution system state estimation (DSSE) in MV systems. Such pseudo load can be represented by the aggregation of smart metering (SM) data. This follows the government restriction that distribution network operators (DNOs) can only use aggregated SM data. Therefore, we assess the subsequent performance of the DSSE, which shows the impact of this restriction - it affects the voltage angle estimation significantly. The possibilities for improving the DSSE accuracy under this restriction are further studied. First, two strategies that can potentially relax this restriction's impact are studied. First, the correlations among pseudo loads' errors are taken into account in the DSSE process and a power loss estimation method is proposed. Second, the investments (i.e., either adding measurement devices or increasing the original devices' accuracy) for the satisfactory DSSE results are assessed. All these are for addressing DNOs' concerns on this restriction.

This paper proposes a Multi-service, User Self-controllable and Privacy-preserving (MUSP) system for secure smart metering. This system has a number of novel properties. Firstly, it can report users' fine-grained consumption data to grid operators and suppliers securely and with user privacy preservation capability. These are achieved by using a homomorphic encryption technique in conjunction with selective data aggregation and distribution methods, so only the aggregated data are delivered to the authorised data recipients only on a need-to-know basis. Secondly, it allows suppliers to access their customers' attributable meter readings regularly. To protect users' privacy, suppliers, by default, can access new data only at a low frequency (e.g. once a month). However, MUSP allows users (1) to adjust (control) this frequency and (2) to release new data by demand (e.g. when change of tariff occurs), thus putting users' privacy preservation in their own hands. Thirdly, it is equipped with an easy and user friendly supplier switching facility to allow users to switch providers easily and conveniently. Security analysis and performance evaluation demonstrate that the MUSP system can protect users' privacy while providing these services in an efficient and scalable manner.

This paper proposes an activity recognition method for streaming sensor data in smart homes. Experiments on real datasets have been carried out to show the effectiveness of the proposed approach. It has great potential for applications in e-health and assisted living.

The IEEE 802.11 Working Group has initiated a new study group known as IEEE 802.11ax which is aiming to devise ways to improve spectrum efficiency, in particular to enhance the system throughput in highly dense scenarios, frequently referred to as Overlapped Basic Service Set (OBSS). In this paper we revisit some of the common problems faced in traditional WiFi networks and show how their effects could be amplified in dense deployments, especially in co-channel scenarios. We then highlight our findings through a simulation based study and draw inferences from these. Some of the key insights from this study are: link suppression and deadlock effects could potentially amplify in co-channel deployments thereby significantly degrading the throughput performance. Also, increasing the concentration of APs in a given area may not always lead to better performance and therefore AP placement needs to be carefully managed in OBSS scenarios. Where AP placement cannot be controlled due to unmanaged environments, findings indicate the need for intelligent load balancing and channel selection algorithms to minimize the impact of the aforementioned effects.

The rapidly increasing demand of M2M (Machine to Machine) communications poses great challenges to the capacity of cellular networks. This paper proposes a new M2M scheduling algorithm, namely, Class Based Overall Priority (CBOP) scheduling, which is designed particularly to improve uplink scheduling for a massive number of MTCDs (Machine Type Communication Devices) in LTE networks. We compare the proposed algorithm with several existing scheduling algorithms via simulations and discuss its advantages and limitations.

This paper discusses malicious false data injection attacks on the wide area measurement and monitoring system in smart grids. First, methods of constructing sparse stealth attacks are developed for two typical scenarios: 1) random attacks in which arbitrary measurements can be compromised; and 2) targeted attacks in which specified state variables are modified. It is already demonstrated that stealth attacks can always exist if the number of compromised measurements exceeds a certain value. In this paper, it is found that random undetectable attacks can be accomplished by modifying only a much smaller number of measurements than this value. It is well known that protecting the system from malicious attacks can be achieved by making a certain subset of measurements immune to attacks. An efficient greedy search algorithm is then proposed to quickly find this subset of measurements to be protected to defend against stealth attacks. It is shown that this greedy algorithm has almost the same performance as the brute-force method, but without the combinatorial complexity. Third, a robust attack detection method is discussed. The detection method is designed based on the robust principal component analysis problem by introducing element-wise constraints. This method is shown to be able to identify the real measurements, as well as attacks even when only partial observations are collected. The simulations are conducted based on IEEE test systems.

Facilitated by advanced information and communication technologies (ICT) infrastructure and optimization techniques, smart grid has the potential to bring significant benefits to the energy consumption management. This paper presents a game theoretic consumption scheduling framework based on the use of mixed integer programming (MIP) to schedule consumption plan for residential consumers. In particular, the optimization framework incorporates integration of locally generated renewable energy in order to minimize dependency on conventional energy and the consumption cost. The game theoretic model is designed to coordinatively manage the scheduling of appliances of consumers. The Nash equilibrium of the game exists and the scheduling optimization converges to an equilibrium where all consumers can benefit from participating in. Simulation results are presented to demonstrate the proposed approach and the benefits of home demand management.

We consider an opportunistic spectrum usage model facilitated with periodic spectrum sensing and handoff in a two-hop selective relay network. A novel sensing and fusion algorithm is proposed under a signalling bandwidth-constrained condition by introducing a quantized soft sensing and a test statistic restoration process. The reliability of secondary transmissions is studied, where we derive expressions for the probability of collision, and the throughput of secondary users. Numerical results manifest the advantages of the proposed sensing algorithm in selective relay networks in light of the missed detection probability, signalling cost, collision probability and throughput.

Primarily the article briefly introduces the basic concept of the flat earth removal and baseline estimation, and mainly introduces the method of flat earth removal and baseline estimation based on orbit parameters using Radarsat-2 image, in detail represents the processing of flat earth removal based on orbit parameters. By removing the flat earth phase of the interferogram and analyzing the result of flat earth removal, the accuracy of generated digital elevation model and surface deformation, it shows that the method is effective, thanks to the highly precise orbit parameters. Meanwhile, the baseline estimation can be operated, which improves the efficiency.

With the deployment of numerous femtocells within macrocells, the rapidly expanding femto base stations have to be self-organizing for interference management, coverage and capacity optimization. This paper focuses on the issue of downlink power control for femtocell self-organization. Different downlink power control strategies have been thoroughly studied with realistic LTE system parameters. We have shown that downlink power control can effectively balance the performance between macrocell and femtocell, and mitigate interference or optimize coverage and capacity. Our study provides useful guidelines for network operators in practical network planning, in which a careful balance should be struck in maintaining satisfactory network performance and user QoS in heterogeneous networks.

In this paper, a new spectrum allocation method is proposed to address the downlink interference problem in heterogeneous macro-femto networks, where distributed femtocells coexist with macrocells. In practical heterogeneous macro-femto networks, normally there is no direct link between a macro base station and a femto base station, and any communication between the two has to be via the backhaul and hence delay-tolerant. Meanwhile, due to wide deployment of distributed femtocells, the signaling overhead via the backhaul should be kept as low as possible. To this end, this paper comes up with a dynamic fractional frequency reuse and dynamic resource allocation scheme to tackle the interference problem, taking into consideration the delay- tolerance and low overhead requirements. Performance evaluation results have demonstrated the benefits of the proposed scheme.

Machine to machine (M2M) communication is the direct communication between devices without human involvements. Recently M2M services in cellular networks have been promoted by cellular operators for sustainable profits for the next decade. However, M2M will also bring serious access congestion problems into cellular networks. This paper studies random access (RA) protocols for cellular M2M communication systems and in particular investigates two RA methods including transmission probability control and two-hop Aloha as well as their dynamic extensions. Our analysis illustrates that the proposed solutions can provide much better performance in terms of random access throughput and delay.

In this paper, the problem of location classification in OFDMA-based FFR systems is considered for two FFR schemes, namely Strict FFR (FFR-A) and Soft Frequency Reuse (FFR-B). The FFR systems considered in the literature are mostly studied with static resource partitioning where a greater portion of the sub-carrier resources are reserved for users in the cell centre and a smaller portion for users near the edge. Once the resources are statically partitioned, they can then be allocated dynamically. In order to ensure proper resource allocation, the base station has to classify the location of the user as either cell-centre or cell-edge. A common practice is to use a one-dimensional threshold such as SINR, which is assumed to be a good indicator of distance. In this paper, the impact of misclassifying the user location on the overall system performance is studied. It is shown that in an urban environment with shadowing, the one-dimensional threshold approach gets the location classification right only 67% of the time and when compared to an ideal system based on accurate location, the overall cell throughput drops by 38% for FFR-A and 14% for FFR-B. Similarly, service rate drops of 24% and 28% are also observed for FFR-A and FFR-B respectively. A better technique based on a combination of two measurements, namely received power and SINR, is proposed where the higherdimensional threshold is determined through Logistic Regression. This new approach is shown to have a classification accuracy of more than 80%. As a result, the system performance is shown to be much better than the one-dimensional threshold approach and comparable to that based on accurate location.

This study exploits coordinated proportional fair (PF) user scheduling for in‐building distributed antenna systems (DAS) to address the co‐channel interference problem. The PF algorithm for multiple remote units in the DAS has been analysed and two sub‐optimal solutions have been proposed, which manage to maximise the subcarrier‐wise summation of PF ratio either in a brute‐force search manner or in a greedy manner with lower complexity. System‐level evaluation shows that the throughput of the cell edge users can indeed be significantly improved, whereas the average throughput of radio units in the proposed schemes also outperforms that in the conventional PF scheme.

ABSTRACT Recently, smaller cells such as femtocells have been proposed to address the cellular coverage problem and provide a ‘greener’ solution to future high‐speed wireless access. On the other hand, interference is a serious problem that could impact the wide deployment of femtocells. The contributions of this paper are threefold. First, we focus on the trade‐off between energy efficiency and system performance. With realistic long‐term evolution system parameters, we have shown that femtocells can indeed improve energy efficiency of the network. However, this comes with a price—performance degradation due to interference, which is especially severe in densely deployed scenarios. Second, we propose a proactive approach to handover and access management in femtocell systems, focusing especially on the interference issue of closed subscriber group femtocells. Third, we propose an efficient data offloading mechanism for interference mitigation and mobility management, with the aim to avoid potential interference and save radio resources and signalling load in the network. Simulation results have been presented to demonstrate the benefits of the proposed schemes. Copyright © 2011 John Wiley & Sons, Ltd.

This paper considers the spectrum management problem in a secondary cognitive radio network where secondary nodes in the network operate on a single channel. There is a leader node which is responsible for tasks such as primary detection, node coordination and channel scanning. Channels are sensed in the order of their detection probabilities. The leader node changes on each scan depending on the credit history generated using a fair credit based mechanism. It has been shown that the proposed scheme can achieve better energy efficiency in terms of overall low battery consumption, better fairness, and seamless switch over when migrating to another channel.

In this paper, we present a novel design named CARLA which combines Cooperative Relaying and Link Adaptation for IEEE 802.11 wireless networks, in order to maximize spectral efficiency and hence improve the throughput of the network. Unlike previous studies assuming that accurate channel information is available for relaying and link adaptation, the novelty of our approach is to use the expected packet transmission time (ETT) and RTS/CTS exchange to estimate the quality and level of contention for the current channel. By taking into account the quality of the direct channel from source to destination and the relay channel between relay and destination, CARLA enables both source and relay terminals to adjust their transmission rates so that not only the reliability of the transmission but also the bandwidth efficiency can be improved, hence fully utilizing the diversity gain. CARLA is a simple, distributed and rate-per-link adaptation protocol requiring no modification on 802.11 PHY and MAC. Simulation results show that CARLA can achieve significant performance improvement in terms of end-to-end throughput and energy efficiency for different network conditions.

IEEE 802.11 Wireless LANs (WLANs) use rate adaptation algorithms (RAAs) to dynamically switch data rates to accommodate the fluctuating wireless channel conditions. Classic RAAs such as ARF and ONOE suffer from rate poisoning and inability to distinguish between collision and packet losses caused by channel errors. Existing approaches in the literature are able to solve only one of the above two issues. In this paper we propose a novel rate adaptation protocol to address both issues for multi-rate wireless networks. The novelty of our approach is to combine the metrics of expected packet transmission time (ETT) and the average number of frozen slots (ANFS) to estimate the quality and level of contention for the current channel. A mathematical model to calculate ETT and ANFS on the fly is presented. Our protocol is simple and practical, which takes into consideration not only link quality and frame loss characteristics, but also impact of collisions during the design. Simulation results show that without the perfect knowledge of current channel condition and any signal strength information, our algorithm can achieve significant performance improvement in terms of end-to-end throughput for different network conditions compared with state-of-the-art link adaptation algorithms.

This paper proposes a distributed framework for demand response and user adaptation in smart grid networks. In particular, we borrow the concept of congestion pricing in Internet traffic control and show that pricing information is very useful to regulate user demand and hence balance network load. User preference is modeled as a willingness to pay parameter which can be seen as an indicator of differential quality of service. Both analysis and simulation results are presented to demonstrate the dynamics and convergence behavior of the algorithm.

One of the fundamental challenges in wireless ad-hoc networking is promotion of cooperation among rational and selfish nodes. Various schemes have been proposed in the literature to stimulate nodes collaboration by improving data forwarding function or route discovery function. The novel algorithm proposed in this paper enhances these two ad-hoc networking functions by utilizing a unique credit-reputation hybrid approach. In the data forwarding phase, a reputation and credit combined mechanism offers incentives with simpler and less overhead process; in the route discovery phase, a new routing metric considering the remaining energy as well as balance between credits and energy has been introduced. Simulations results show that the scheme could successfully stimulate node cooperation by manipulating credits and reputation restrictions and more importantly actively discouraging three types of selfish behaviors.

In this paper we propose an efficient operation mechanism of interference mitigation and mobility management. The main idea is to schedule radio resources between macrocell and femtocell base stations and pass the desired packet data through a femtocell base station without handover/access to the CSG (closed subscriber group) femtocell. The key benefit of this proposed mechanism is to achieve highly efficient mobility by enabling a CSG femtocell base station to relay data without handover and access to the CSG femtocell, avoiding potential interference and saving radio resource and signaling load on the network. Simulation results have been presented to demonstrate the benefits of the proposed scheme.

In this paper, we study power loading and resource allocation for OFDM-based femtocell networks. To mitigate the interference to the macrocell from the femtocell, interference constraints have to be applied for power loading in the femtocell. Taking the interference constraints into account, we propose an iterative water filling algorithm to improve the system performance. The proposed algorithm is further developed with a proportional fair scheduling algorithm to maintain the fairness among the femto users. Simulation results are given to support the advantages of the proposed solution, showing that more than 60% capacity gain at -8dB SNR and 5% gain at 6dB can be achieved.

Motivated by the need to improve energy efficiency and reduce the impact on the environment, we see efforts directed towards enabling a `Smart Grid'. One of the key elements of the Smart Grid is the Automated Metering Infrastructure (AMI) which is expected to facilitate the transport of meter readings from meters to the utility provider and control information (potentially) in the other direction. Different solution strategies for realizing AMI networks are currently being debated in the community with no clear consensus so far. This paper proposes a mesh radio based solution which exhibits self-organizing characteristics and is therefore attractive from a deployment perspective. The proposed solution has been realized as enhancements to the RPL protocol, a connectivity enabling mechanism for low power and lossy networks currently being standardized by the IETF ROLL working group. It enables smart meter nodes augmented with the mesh radio to automatically discover concentrator nodes in the vicinity, setup a single/multi-hop link to the best available concentrator, detect loss of connectivity and subsequently re-configure itself to re-establish connectivity so as to ensure reliable transport of smart metering data. Moreover, it achieves this in a distributed manner thereby ensuring scalability. Results from the evaluation of the proposed solution through emulation based experiments further confirm this.

In cognitive radio networks, the statistical duty cycles (traffic loads) of primary channels have a decisive effect on the throughput and link maintenance of secondary users (SUs). It is very difficult and time-consuming for the SUs to distinguish between primary channels in terms of duty cycles. This paper proposes a multi-user cooperation scheme where SUs share status information about the primary channels and the corresponding cooperation information processing scheme based on Bayesian inference to differentiate the primary channels in terms of statistical duty cycles within a short period of time. The cooperation information processing scheme helps the SUs quickly make accurate estimates of the statistical duty cycles of primary channels and theire uncertainties. Specifically, with the aid and information from its cooperating neighbors, a SU will continuously update the likelihood function and the priori distribution for each primary channel. Analytical and simulation results show that the proposed multi-user cooperation scheme, which is based on a Bayesian updating rule, can significantly save a SU's time in channel detection and selection, and therefore greatly improve its data transmission time and throughput.

Link adaptation is a critical component of IEEE 802.11 systems, which adapts transmission rates to dynamic wireless channel conditions. In this paper we investigate a general cross-layer link adaptation algorithm which jointly considers the physical layer link quality and random channel access at the MAC layer. An analytic model is proposed for the link adaptation algorithm. The underlying wireless channel is modeled with a multiple state discrete time Markov chain. Compared with the pure link quality based link adaptation algorithm, the proposed cross-layer algorithm can achieve considerable performance gains of up to 20%.

In this paper we discuss the application of multiple-wireless technology to a practical context-enhanced service system called ViewNet. ViewNet develops technologies to support enhanced coordination and cooperation between operation teams in the emergency services and the police. Distributed localisation of users and mapping of environments implemented over a secure wireless network enables teams of operatives to search and map an incident area rapidly and in full coordination with each other and with a control centre. Sensing is based on fusing absolute positioning systems (UWB and GPS) with relative localisation and mapping from on-body or hand-held vision and inertial sensors. This paper focuses on the case for multiple-wireless capabilities in such a system and the benefits it can provide. We describe our work of developing a software API to support both WLAN and TETRA in ViewNet. It also provides a basis for incorporating future wireless technologies into ViewNet.

In this paper a simple and effective method has been proposed to perform access and handover management for femtocell systems. Firstly an analysis on interference impact on femtocells has been provided. Then we explore several techniques to address the interference problem between CSG (closed subscriber group) femtocell and macrocell. When a non-CSG MS comes close to the femtocell in question, the macrocell BS makes the decision (based on the CSG access list exchanged from the femtocell) that unnecessary handover will not be initiated, saving radio resource and signaling load on the network. At the same time various proactive interference mitigation mechanisms can be operated. A hybrid access mode and a femtocell-initiated handover procedure with adaptive threshold have also been discussed.

Radio frequency identification (RFID) technology has been gaining growing interest from a wide range of industries because of its lower cost and increased capability. In this paper, we study the performance of RFID localization in an indoor environment. In particular, we investigate the k-Nearest Neighbour (k-NN) algorithm in different indoor radio propagation environments and different deployment scenarios and settings. By modelling these various indoor features, the k-NN algorithm is tested and observed to be robust in terms of the estimation errors in various indoor localization scenarios. On the other hand, this study reveals that its performance can be improved by tuning certain parameters, e.g. the k value and different weighting methods adaptively for different locations.

In future dynamic spectrum access (DSA) wireless networks, secondary users can make use of spectrum opportunities dynamically for communications. These opportunities are available when primary users do not use their licensed spectrum slots. In this paper we study the scenario where spectrum opportunities are characterized by the ON/OFF distribution. As a result, the queueing system can be analyzed using the well-established theory of fluid flow models and effective bandwidth. Based on a refinement to the original effective bandwidth approximation, we derive the packet loss probability for secondary users access to the available channels. Theoretical and numerical analyses are used to investigate the impacts of various parameters of the underlying DSA link model on the system performance.

Underutilization of the limited spectrum sparked the need for dynamic spectrum access and flexible spectrum policies. Accurate estimation of spectrum occupancy is an essential step for spectrum access in distributed networks. This paper analyzes a bidirectional and dual scanning method that scans multiple spectrum bands to find a suitable free-channel. Moreover, secondary users' access to available spectrum could be interrupted by the arrival of primary spectrum users, forcing the well-behaved secondary users to switch to another free-channel. A pointer and adaptive reserve channel mechanism is proposed in this paper to facilitate quicker and smoother channel switching. Simulation results for both techniques show improved overall connectivity is maintained for varying primary users' arrival rates.

Because of the short-range characteristics of UWB or 802.15.3 networks (around 10 m), it is highly desirable to have a range extension technology for future wireless personal area networks (WPANs). In the 802.15.3 standard, coverage extension is achieved by creating child or neighbor piconets from a parent piconet. One of the main disadvantages of this approach is that it does not make use of the multi-channel capability provided by the underlying physical layer (e.g. multi-band OFDM) as both the dependent and parent piconets operate on the same channel. As a result the capacity of the wireless medium is not fully utilized and the service quality of high rate A/V applications will suffer. This paper first describes a device association scheme to avoid unreachable communication pairs in network initialization. It then proposes a mesh network architecture and a simple, low-cost method for establishing communications across multiple piconets via multi-hop connections. It can be used to extend network coverage through connecting piconets on multiple channels.

Medium access control (MAC) protocols have a large impact on the achievable system performance for wireless ad hoc networks. Because of the limitations of existing analytical models for ad hoc networks, many researchers have opted to study the impact of MAC protocols via discreteevent simulations. However, as the network scenarios, traffic patterns and physical layer techniques may change significantly, simulation alone is not efficient to get insights into the impacts of MAC protocols on system performance. In this paper, we analyze the performance of IEEE 802.11 distributed coordination function (DCF) in multihop network scenario. We are particularly interested in understanding how physical layer techniques may affect the MAC protocol performance. For this purpose, the features of interference range is studied and taken into account of the analytical model. Simulations with OPNET show the effectiveness of the proposed analytical approach.

A service discovery protocol for ad hoc networks based on the IETF service location protocol (SLP) has been implemented in ns. Its performance is studied under different mobility patterns and number of service agents. A lightweight quality of service (QoS) aware service discovery protocol is also proposed to support QoS applications.

A wireless ad hoc network is an autonomous system where all nodes are capable of movement and can be connected dynamically in an arbitrary manner. Without a network infrastructure, network nodes function as routers which discover and maintain routes to other nodes in the network. Routing in ad hoc networks has received a lot of attention due to the challenges posed by the self-organizing nature of the network, the dynamic topology, and the unreliable wireless medium. Quality of service (QoS) routing searches for routes with sufficient resources for the QoS requirements and is one of the core technologies to support QoS for real-time multimedia and data applications in ad hoc networks. We show that MAC layer information can be utilized at the routing layer via certain routing metrics for optimal route selection. We propose to use MAC delay as an indication of the level of interference of a link. Link reliability and residual bandwidth are two other routing constraints. Then, the resulting NP-complete multiple constraint routing problem is solved using a heuristic algorithm. The performance of the algorithm is studied using simulations.