David J. Thuente

Improving quality of service for MIL-STD-188-220C

Quality of service (QoS) is an important component for the military communication networks. QoS for combat network radios can be critical for the efficiency and, indeed, the survival of a mission and its personnel. The implementation of QoS for the combat network radio networks and the tactical Internet is largely restricted to the utilization of priority (generally referred to as precedence in the fire support (FS) community). QoS for MIL-STD-188-220C is generally implemented in the data link layer using media access control (MAC) algorithms. A legitimate criticism of MIL-STD-188-220B/C in terms of implementing QoS is its restriction to three levels of precedence. An important consequence of the current implementation of QoS for the most urgent and time critical messages such as check-fires is that they are just another urgent message in terms of network access for all MIL-STD-188-220C MAC algorithms. The time criticality of these types of messages is obvious but even if there is a near universal requirement not to use urgent precedence messages, with current MAC implementations, urgent does not adequately provide the QoS needed for the almost immediate delivery of such messages. This paper presents two techniques for significantly improving the QoS performance for these most time critical messages. One additional level of precedence can be added to MIL-STD188-220C that improves the performance at all precedence levels. The study assumes a benign fully connected network. The paper discusses how the methodology used in this study could be applied to other communication systems.

Jamming attacks in 802.11g — A cognitive radio based approach

Wireless networks are susceptible to jamming attacks, which can severely reduce the network throughput. In this paper, we study the behavior and the performance of 802.11g networks under a hybrid jamming attack of configuring a cognitive radio as a jammer. With characteristics such as fast channel switching, quick response time and software reconfigurability, cognitive radios can be used not only to improve the spectrum sharing management, but also to act as an effective jammer. We use a single cognitive radio to simultaneously jam three networks in an energy efficient manner and also to deny any channel change protocol by the targeted network to avoid jamming. We attack the ‘g’ band OFDM channels directly using the fast channel switching capability of the cognitive radio. The jammer sequentially senses traffic on each of the networks without being part of any network. We present the results of the jamming attacks at the MAC and physical layers. We show how the cognitive radio can dynamically adjust its attack to the traffic on each network. We evaluated the performance of three networks individually and together under intelligent and reactive jamming.

Channel Detecting Jamming Attacks against Jump-Stay Based Channel Hopping Rendezvous Algorithms for Cognitive Radio Networks

Recently many channel hopping algorithms have been studied to guarantee rendezvous for Cognitive Radio Networks (CRNs). These algorithms propose rendezvous methods without using common control channels (CCCs) to avoid the limitations such as single point of failure, low scalability, and jamming attacks. In particular, the Jump-Stay based channel hopping rendezvous (JSR) algorithms provide guaranteed rendezvous for CRNs with no time synchronization or CCCs (i.e., blind rendezvous). However, the JSR algorithms are still vulnerable to Channel Detecting Jamming Attacks (CDJAs) in which the jammer can estimate the channel hopping sequences within the first jump-pattern. The jammer can compute the entire JSR channel hopping sequence and thus reduce the rendezvous success rate from 100% to less than 20% and 10% using one and two listening channels respectively. To mitigate this problem, we revisit both the Random rendezvous scheme and the Role-based Channel Rendezvous (RCR) scheme extended from role- based rendezvous algorithms to increase the probability of the rendezvous against the CDJAs. We also compare the JSR algorithm to both the Random and RCR algorithms and show the Random and RCR vastly outperform the JSR algorithm when there are security concerns about a channel detecting jammer. Especially, the effectiveness of CDJA is negligible for the Random and RCR schemes but their expected time to rendezvous (TTR) is close to the JSRs expected TTR.

Limitations of Quorum-based Rendezvous and key establishment schemes against sophisticated jamming attacks

Recently Quorum-based frequency hopping schemes have been studied to increase rendezvous probabilities and to provide fast key establishment techniques in RF communication under jamming attacks. However, these schemes are still vulnerable to sophisticated jamming attacks in which a jammer has the capability of listening and jamming multiple frequencies. In this paper, we present a sophisticated jamming attack and evaluate its effectiveness on Frequency Quorum-based Rendezvous (FQR) schemes [4]. The sophisticated jammer can find the senders quorum set within the second frame (i.e., 2fc time slots) in the FQR system when it has the capability of listening on k frequencies from the minimal (N, k) difference sets. The jammer can completely jam the sender after 2fc time slots using an average of ⌊k+1/2⌋ and a maximum of k frequencies. To remedy this jamming problem, we revisit the role-based rendezvous scheme and extend it to the Role-based Frequency Rendezvous (RFR) scheme. Our simulation results demonstrate that the rendezvous probability of the FQR system under the sophisticated jamming attack dramatically decreased as the number of available channel N increases (e.g., ≤ 35% for N ≥ 30 in k2 time slots). On the other hand, the rendezvous probability of our RFR scheme is almost steady for all available frequencies N (e.g., ≥ 90% for N ≥ 20). Therefore, the RFR scheme can be an effective, efficient and robust rendezvous and key establishment scheme against sophisticated jamming attacks.

Integrated simulation and emulation using adaptive time dilation

Simulation and emulation techniques are commonly used to evaluate the performance of complex networked systems. Simulation conveniently predicts the behavior of a complex networked system while usually requiring fewer simplifying model assumptions often necessary for theoretical analysis. In contrast, emulation does not need to re-implement the target real systems, so it may improve on the implementation efficiency of simulation while maintaining much of the realism of testbeds. A hybrid approach in which simulation nodes connect to emulation hosts can be used to combine the advantages of both approaches. In this paper, we propose integrating simulation with emulation using adaptive time dilation to evaluate system performance. If a simulator schedules its events in real time and the simulation time keeps up with the real time, then the hybrid system works very well and meets its deadlines. However, a heavily-loaded simulator can introduce significant simulation delays and thereby create situations where these delays impact the accuracy of the system. Our approach uses time dilation to reduce simulation delays and thus increasing the accuracy of the integrated simulation and emulation system. Our adaptive time dilation dynamically controls the time dilation factor to avoid system overloads for both the simulation and the emulation components and to improve the execution correctness of the hybrid system.

Sequence sensing Jamming attacks against modular-based channel hopping rendezvous algorithms for cognitive ratio networks

Efficient utilization of wireless bandwidth is a critical, if not the key, component of the wireless network architectures that balance availability and access. Cognitive Radio Networks (CRNs) are an important part of the solution to this problem. Common control channels (CCCs) for rendezvous in CRNs have limitations such as single point of failure, low scalability, and susceptibility to jamming attacks. Several rendezvous algorithms have recently been proposed that remove the need for CCCs. In particular, the Modular Clock (MC) channel hopping rendezvous algorithms provide extremely efficient rendezvous times for CRNs without using time synchronization and CCCs (i.e., blind rendezvous). However, these algorithms are vulnerable to Sequence Sensing Jamming Attacks (SSJAs) in which the jammer can estimate the channel hopping sequences within the first-half period of the MC algorithm. Using a single jammer and two listening channels, we show how to compute the entire MC Channel Hopping (CH) sequence and thus reduce the rendezvous success rate from around 95% to around 15% for the basic period. We show this is a major security problem for CRNs utilizing the MC algorithm since any secondary user or even group of users can easily be denied access to the network with high probability. We also compare these results to the Random rendezvous algorithm and show it vastly outperforms the MC algorithm when there are security concerns about a sequence sensing jammer.

Using Decision Trees for State Evaluation in General Game Playing

A general game playing agent understands the formal descriptions of an arbitrary game in the multi-agent environment and learns to play the given games without human intervention. In this paper, we present an agent that automatically extracts common features shared by the game winners and uses such learned features to build decision trees to guide the heuristic search. We present data to show the significant performance improvements contributed by the decision tree evaluation. We also show by using hash tables in knowledge reasoning, our agent uses 80% less time when compared to a widely available GGP agent written in the same language.

Dynamic Localization Schemes in Malicious Sensor Networks

Wireless sensor networks (WSN) have recently shown many potential military and civilian applications, especially those used in hostile environments where malicious adversaries can be present. The accuracy of location information is critical for such applications. It is impractical to have a GPS device on each sensor in WSN due to costs. Most of the existing location discovery schemes can only be used in the trusted environment. Recent research has addressed security issues in sensor network localization, but to the best of our knowledge, none have completely solved the secure localization problem. In this paper, we propose novel schemes for secure dynamic localization in sensor networks. These proposed schemes can tolerate up to 50% of beacon nodes being malicious, and they have linear computation time with respect to the number of reference nodes. Our security analysis has showed that our schemes are applicable and resilient to attacks from adversaries. We have further conducted simulations to analyze and compare the performance of these schemes, and to indicate when each should be used. The efficiencies of each method shows why we needed to propose multiple methods.

Secure localization schemes in sensor networks

The accuracy of location information is critical for many applications of wireless sensor networks (WSN), especially those used in hostile environments where malicious adversaries can be present. Since it is impractical to have a GPS device on each sensor in WSN, most of the existing location discovery schemes can only be used in the trusted environment. Recently, some advanced research has addressed security issues in sensor network localization, but to the best of our knowledge, none of them has completely solved the secure localization problem. In this paper, we propose novel schemes for secure localization in sensor networks. One of these proposed schemes requires only linear computation time with respect to the number of reference nodes. Our security analysis has shown that our schemes are applicable and resilient to attacks from adversaries. We have further conducted simulations to analyze the performance of these schemes.

High-performance emulation of heterogeneous systems using adaptive time dilation

Building a testbed for evaluating the performance of large-scale heterogeneous systems can be costly and inefficient. Emulation is often used to evaluate the performance of a system in a controlled environment. Time dilation allows virtual machines (VMs) to emulate higher performance than that of their physical machine. We present an approach using adaptive time dilation to emulate large-scale distributed systems composed of heterogeneous machines and Operating Systems (OSs). In our implementation, VMs are globally synchronized. To evaluate our system, distributed VMs running Linux, Windows, FreeBSD, and Junos are emulated on general-purpose physical machines.