The increased deployment of ubiquitous wireless networks has exponentially increased the complexity to detect wireless network attacks and protect against them. In this research, we investigated the vulnerabilities in wireless networks, developed a comprehensive taxonomy of wireless attacks that has been used to guide our approachto develop, and successfully implement a self-protection wireless system capable of detecting and protecting wireless networks from a wide range of attacks.In the past few years, more security improvements took place, but the network is still vulnerable to complex, dynamic, and knowledgeable attacks; in addition, a large number of last-generation unsecured network cards are still available on the market. This dissertation presents an anomaly-based wireless intrusion detection and response system, which is capable of detecting complex malicious attacks. Our approach is based on multi-channel online monitoring and analysis of wireless network features with respect to multiple observation time windows. These features are related to Data Link Layer framebehaviors and the mobility of stations. We have successfully designed and implemented A Wireless Self Protection System (WSPS) that has the following significant features: it monitors wireless networks, generates network features, tracks wireless-network-state machine violations, generates wireless network flows (WNetFlows) for multiple time windows, and uses the dynamically updated anomaly and misuse rules to detect complex known and unknown wireless attacks and take appropriate proactive actions. To evaluate the performance of WSPS and compare it with other wireless intrusion detection systems, we present an evaluation approach that uses different metrics such as adaptability, scalability, accuracy, overhead, and latency.We validate the WSPS approach by experimenting with normal traffic and more than 20 different types of wireless attacks; and compare the WSPS performance with several well-known intrusion protection systems. Our experimental results show that the WSPS approach can protect from wireless network attacks with an average detection rate of 99.13% for all the experimented attacks.

During the route discovery process, each node receiving the route request packet (RReq) will retransmit it exactly once. A distant neighbor may accidentally receive/loose the only RReq and use it to announce a new route, although that link is inferior/superior for route reply packets (RRep) or actual message routing. Overall, the constructed route may be far from the optimal. All existing route discovery schemes (including DSR/AODV) apply retransmission during route discovery exactly once (1R). Based on a realistic physical layer model, we propose two new route discovery schemes: n-retransmission (nR, retransmitting exactly n times) and n-retransmission c-reception (ncRR), retransmitting until we either reach a total of n own retransmissions or c copies from neighbors are heard. We compare our two new scheme with the traditional one, under otherwise identical conditions (same metric, same packet reception probability on each link) and the same choices about possibly retransmitting again upon discovering a better route (R+) or discarding it (R1), generating route reply packet for every received RRep (B*), or for first and better discovered routes only (B2), and retransmitting RRep exactly once (A1), up to a maximum of three times (A3), or optimally u times decided by link quality (Au). Experimental results show that the proposed ncRR scheme (for n=2 and c=3 or c=4) achieves the best tradeoff between quality of route, success rate and message overhead in the route discovery process, followed by the nR scheme, and both of them are superior to the existing traditional schemes.

During the route discovery process, each node receiving the route request packet (RReq) will retransmit it exactly once. A distant neighbor may accidentally receive/loose the only RReq and use it to announce a new route, although that link is inferior/superior for route reply packets (RRep) or actual message routing. Overall, the constructed route may be far from the optimal. All existing route discovery schemes (including DSR/AODV) apply retransmission during route discovery exactly once (1R). Based on a realistic physical layer model, we propose two new route discovery schemes: n-retransmission (nR, retransmitting exactly n times) and n-retransmission c-reception (ncRR), retransmitting until we either reach a total of n own retransmissions or c copies from neighbors are heard. We compare our two new scheme with the traditional one, under otherwise identical conditions (same metric, same packet reception probability on each link) and the same choices about possibly retransmitting again upon discovering a better route (R+) or discarding it (R1), generating route reply packet for every received RRep (B*), or for first and better discovered routes only (B2), and retransmitting RRep exactly once (A1), up to a maximum of three times (A3), or optimally u times decided by link quality (Au). Experimental results show that the proposed ncRR scheme (for n=2 and c=3 or c=4) achieves the best tradeoff between quality of route, success rate and message overhead in the route discovery process, followed by the nR scheme, and both of them are superior to the existing traditional schemes.

Wireless network analysis and simulations rely on accurate physical layer models. The increased interest in wireless network design and cross-layer design require an accurate and efficient physical layer model especially when a large number of nodes are to be studied and building the real network is not possible. For analysis of upper layer characteristics, a simplified physical layer model has to be chosen to model the physical layer. In this dissertation, the widely used two-state Markov model is examined and shown to be deficient for low to moderate signal-to-noise ratios. The physical layer statistics are investigated, and the run length distributions of the good and bad frames are demonstrated to be the key statistics for accurate physical layer modeling. A four-state Markov model is proposed for the flat Rayleigh fading channel by approximating the run length distributions with a mixture of exponential distributions. The transition probabilities in the four-state Markov model can be established analytically without having to run extensive physical layer simulations, which are required for the two-state Markov model. Physical layer good and bad run length distributions are compared and it is shown that the four-state Markov model reasonably approximates the run length distributions. Ns2 simulations are performed and the four-state Markov model provides a much more realistic approximation compared to the popular two-state Markov model. Achieving good results with the flat Rayleigh fading channel, the proposed four-state Markov model is applied to a few diversity channels. A coded orthogonal fre- quency division multiplexing (OFDM) system with a frequency selective channel and the Alamouti multiple-input multiple-output system are chosen to verify the accuracy of the four-state Markov model. The network simulation results show that the four-state Markov model approximates the physical layer with diversity channel well whereas the traditional two-state Markov model estimates the network throughput poorly. The success of adapting the four-state Markov model to the diversity channel also shows the flexibility of adapting the four-state Markov model to various channel conditions.

Data integrity and privacy are the two most important security requirements in wireless communication. Most mechanisms rely on pre-share key data encryption to prevent unauthorized users from accessing confidential information. However, a fixed secret key is vulnerable to cracking by capturing sufficient packets or launching a dictionary attack. In this research, a dynamic re-keying encryption protocol was developed to enhance the security protection for IEEE 802.11 and hybrid wireless network. This protocol automatically updates the secret key during the end-to-end transmission between wireless devices to protect the network and the communication privacy. In addition, security analyses are given to verify the protection of this protocol. Experiment results also validate that the dynamic encryption approach can perform as efficiently as other security architectures while providing an additional layer of data protection.

IEEE 802.11 MAC protocol uses two techniques to avoid interference ¡V physical carrier sensing and RTS/CTS handshake (virtual carrier sensing). Ideally, these two techniques can solve most of interference problem. But in normal situation a host cannot receive packet successfully unless the power of receiving signal is much higher than that of interference signal. This fact leads that other host can interfere the reception in farther distance than the distance between the transmitter and the receiver. When the distance between the transmitter and the receiver is larger than certain value, some hosts that are out of the range covered by RTS/CTS packet may still interfere the receiver. This problem can be solved by using a very large physical carrier sensing range, which is limited by antenna and will reduce the network throughput. So it¡¦s not suitable to avoid large interference problem. So far some schemes were proposed to solve or improve this problem, but most of them cause power waste at the same time. In this thesis, we will propose a scheme to look after both interference and power consumption. Simulation results verify that our scheme can solve most interference caused by large interference range with less power consumption.

With rapidly development of wireless network technology in the past few years, hand-held devices can connect to Internet everywhere conveniently by equipped with wireless card. IEEE 802.11 defined two types of wireless networks: infrastructure networks and ad hoc networks. Since ad hoc networks are a kind of networks without access points, efficient routing is an important issue for these networks. A more efficient routing approach involves dividing the network into several clusters by using a clustering algorithm. The purpose of this paper is to investigate a hybrid cluster routing algorithm in mobile ad hoc networks. This investigation proposed a double-manager K-hop clustering algorithm based on the link evaluation. This algorithm not only can elect a more suitable node as the cluster head but also can reduce the overhead of the managers. Furthermore, this study devised maintenance and backup mechanisms to stabilize network performance in the face of network topology changes. The simulation results demonstrate that our proposed algorithm outperforms other clustering algorithm in terms of number of clusters, routing overhead of managers and cluster stability.

In periodic wireless broadcasting, air behaves like a storage medium requiring new data organization and access methods.Due to power limit for the portable units (ex. the palmtop), how to design an energy-saving organization is a key issue.Imielinski et al. have proposed the hashing based schemes, including the Hashing A and Hashing B schemes, to save energy in the progress of getting data of interest. The Hashing B scheme improves the directory miss phenomenon in the Hashing A scheme, where the directory miss is that the client's initial probe comes before the bucket containing his key but after the bucket which contains a proper offset. However, based on these two schemes, if the differences between the minimum overflow and the other overflows are large extremely or the small overflows appear near the rear part of the broadcast file, both schemes have a poor performance. Therefore, in this thesis, we propose four multiple-hashing-functions-based schemes, including the FirstR, FirstL, AvgK and TopK schemes, to overcome such the situations. The basic idea is to use cutlines to divide the broadcast file with N logical buckets into several regions, and then each region may have the different minimum overflow. Since the minimum overflow in each region can be different, we can have different hashing functions for those regions to determine the positions of the designated buckets. Among the proposed schemes, the difference is how to determine the positions of the cutlines. The FirstR scheme finds those cutlines from the right end to the left whenever the difference of overflows of two adjacent logical buckets is greater than or equal to 1. The FirstL scheme finds those cutlines from the left end to the right whenever the difference of overflows of two adjacent logical buckets is greater than or equal to 1. In the AvgK scheme, we first calculate AvgD, the average of the differences of two consecutive overflows whose values are large than or equal to 1. Then we find cutlines from the left end to the right whenever the difference of two adjacent logical buckets is greater than or equal to AvgD. The TopK determines the cutlines by considering the descending order of the differences of overflows. From our performance analysis and simulation study, the performance of the TopK scheme is the best among the proposed schemes. Therefore, we then make a comparison between the TopK scheme and the Hashing B scheme. Since the number of the hashing functions in the TopK scheme is larger than those in the Hashing B, the physical bucket in the TopK scheme is somewhat bigger than that in the m Hashing B scheme. In our simulation, we have considered this factor as well. From our performance analysis and simulation study, we show that the performance of the TopK scheme performs better than that of the Hashing B scheme, even though the above factor about the storage size is considered. The TopK scheme improves the directory miss in the Hashing B scheme; therefore, the average access time is improved excellently.

This study explores the medical services based on the ICT technologies such as the wireless network and positioning. The related history of internet and recently developed technologies of localization will be introduced. The literature review and case study approaches will be applied to explore the medical services using high technologies. Combining new technologies and medical industry requires an integrated platform. How to design such a platform and what functionalities required in that are investigated in this thesis. Our data is mainly from Industrial Technology Research Institute (ITRI), including the platform for homecare and eldercare. Moreover, we also study the product samples available on numerous web sites. The services provided in Taiwan are inspected as well. Due to the increasing aging population worldwide, the medical services incorporating high-tech such as homecare and eldercare have attracted interests from both academics and industry. Once a well developed platform can be integrated, there are many advantages can be expected such as saving the medical resource, enhancing the service quality, and so on. Therefore, the government puts emphasis on such medical integrated scheme such as the U-care project in order to meet the future requirements. We will report the platform constructed by ITRI and analyze the problems encountered in the advanced country such as Japan and America. Finally, how the technologies are developed from the engineer perspective will be clarified. The networking and positioning services from Intel and Microsoft are studied in this work and the evolvements are also illustrated

In this thesis, the features of electric field signatures of narrow bipolar pulses (NBPs) generated by cloud flashes are investigated and their effects on wireless communication systems are studied. A handful amount of NBPs (14.5%) have been observed to occur as part of cloud-to-ground flashes in South Malaysia. Occurrence of NBPs in Sweden has been reported for the first time in this thesis. The electric field waveform characteristics of NBPs as part of cloud-to-ground flashes were similar to isolated NBPs found in Sweden and South Malaysia and also to those isolated NBPs reported by previous studies from various geographical areas. This is a strong indication that their breakdown mechanisms are similar at any latitudes regardless of geographical areas. A comparative study on the occurrence of NBPs and other forms of lightning flashes across various geographical areas ranging from northern regions to the tropics is presented. As the latitude decreased from Uppsala, Sweden (59.8°N) to South Malaysia (1.5°N), the percentage of NBP emissions relative to the total number of lightning flashes increased significantly from 0.13% to 12%. Occurrences of positive NBPs were more common than negative NBPs at all observed latitudes. However, as latitudes decreased, the negative NBP emissions increased significantly from 20% (Sweden) to 45% (South Malaysia). Factors involving mixed-phase region elevations and vertical extents of thundercloud tops are invoked to explain the observed results. These factors are fundamentally latitude dependent. In this thesis, the interaction between microwave radiations emitted by cloud-to-ground and cloud flashes events and bits transmission in wireless communication networks are also presented. To the best of our knowledge, this is the first time such effects are investigated in the literature. Narrow bipolar pulses were found to be the strongest source of interference that interfered with the bits transmission.