Inhye Kang

An efficient state space generation for analysis of real-time systems

State explosion is a well-known problem that impedes analysis and testing based on state-space exploration. This problem is particularly serious in real-time systems because unbounded time values cause the state space to be infinite. In this paper, we present an algorithm that produces a compact representation of reachable state space of a real-time system. The algorithm yields a small state space, but still retains enough timing information for analysis. To avoid the state explosion which can be caused by simply adding time values to states, our algorithm first uses history equivalence and transition bisimulation to collapse states into equivalent classes. In this approach, equivalent states have identical observable events although transitions into the states may happen at different times. The algorithm then augments the resultant state space with timing relations that describe time distances between transition executions. For example, the relation @(tr1) + 3 ≤ @(tr2) ≤ @(tr1) + 5 means that transition tr2 is taken 3 to 5 time units before transition tr2 is taken. This is used to analyze timing properties such as minimum and maximum time distances between events. To show the effectiveness of our algorithm, we have implemented the algorithm and are currently comparing it to other existing techniques which generate state space for real-time systems.

Real-time visualization of network attacks on high-speed links

This article shows that malicious traffic flows such as denial-of-service attacks and various scanning activities can be visualized in an intuitive manner. A simple but novel idea of plotting a packet using its source IP address, destination IP address, and the destination port in a 3-dimensional space graphically reveals ongoing attacks. Leveraging this property, combined with the fact that only three header fields per each packet need to be examined, a fast attack detection and classification algorithm can be devised.

100+ VoIP Calls on 802.11b: The Power of Combining Voice Frame Aggregation and Uplink-Downlink Bandwidth Control in Wireless LANs

The bandwidth efficiency of voice over IP (VoIP) traffic on the IEEE 802.11 WLAN is notoriously low. VoIP over 802.11 incurs high bandwidth cost for voice frame packetization and MAC/PHY framing, which is aggravated by channel access overhead. For instance, 10 calls with the G.729 codec can barely be supported on 802.11b with acceptable QoS - less than 2% efficiency. As WLANs and VoIP services become increasingly widespread, this inefficiency must be overcome. This paper proposes a solution that boosts the efficiency high enough to support a significantly larger number of calls than existing schemes, with fair call quality. The solution comes in two parts: adaptive frame aggregation and uplink/downlink bandwidth equalization. The former reduces the absolute number of MAC frames according to the link congestion level, and the latter balances the bandwidth usage between the access point (AP) and wireless stations. When used in combination, they yield superior performance, for instance, supporting more than 100 VoIP calls over an IEEE 802.11b link. The authors demonstrate the performance of the proposed approach through extensive simulation, and validate the simulation through analysis.

Preventing session table explosion in packet inspection computers

We first show that various network attacks can cause fatal inflation of dynamic memory usage on packet processing computers. Considering Transmission control protocol (TCP) is utilized by most of these attacks as well as legitimate traffic, we propose a parsimonious memory management guideline based on the design of the TCP and the analysis of real-life Internet traces. In particular, we demonstrate that, for all practical purposes, one should not allocate memory for an embryonic TCP connection with roughly more than 10 seconds of inactivity.

Resolving the Unfairness of Distributed Rate Control in the IEEE WAVE Safety Messaging

In the IEEE Wireless Access in Vehicular Environment (WAVE), the periodic broadcast of the basic safety message (BSM) enables proximity awareness in the vehicle-to-vehicle (V2V) context. To maximize the level of awareness and consequently the driving safety, the BSM transmission at the highest allowed rate is desired in principle. A caveat, however, is controlling the BSM traffic within the given channel capacity because otherwise it can actually lower the delivery probability due to message collisions. To avoid such a congestion situation, a traditional mode of control is regulating the frequency of the BSM transmission based on the channel load. In this paper, we shed light on the pitfalls that lurk in exercising adaptive rate control based on an observed global state such as the channel load. Specifically, we show that straightforward threshold- or hysteresis-based controls can irrevocably render the rate assignments irrelevant to the given vehicle density pattern. As a solution, we show that distributed but coordinated control provably leads to stability and relevance to the given vehicle density pattern.

Hop-by-Hop Frame Aggregation for VoIP on Multi-Hop Wireless Networks

The already severe inefficiency problem in running VoIP on the IEEE 802.11-based wireless LAN is exacerbated in multi-hop networks due to spatial interference. This paper presents a novel scheme that significantly mitigates the effect of the interference by combining the inter-call aggregation and the pseudo-broadcast technique as used in network coding. As such, the proposed scheme works effectively not just for calls traveling the same routed path but also for calls crossing each other inside the multi-hop networks. We demonstrate through extensive simulation that the IEEE 802.11-based multi-hop network can be boosted up to 700% in terms of the number of supportable VoIP calls, with acceptable QoS, through the proposed scheme.

Timed and Resource-Oriented Statecharts for Embedded Software

Embedded software should be correctly developed so that it is be compliant with not only functional requirements but also real-time and resource constraints. However, those constraints are often dependent on execution environments that are sometimes revealed in late development phases. In this paper, we propose Timed and Resource-oriented Statecharts (TRoS) to analyze the time and resource-constrained behavior of system in earlier development phases of embedded software development. TRoS extends Statecharts using timed action labeled with resources to represent actions that consume resources. This enables us to describe the competition among processes to use shared resources, and to analyze schedulability of embedded systems. We present a case study of a distance control module that controls train movement to keep the distance between trains for railway control systems.

WLC34-1: A Simple Congestion-Resilient Link Adaptation Algorithm for IEEE 802.11 WLANs

Algorithmic approach to link adaptation for IEEE 802.11 networks such as Automatic Rate Fallback (ARF) is known to suffer from the inability to differentiate between collision and channel-induced error. In this paper, we propose a novel algorithm called COLA that overcomes the shortcoming and achieves near-optimal throughput over wide range of channel and load conditions. The result is significant since the throughput is achieved without any hardware support. Moreover, COLA does not require any optional or extra-protocol mechanisms support, either, such as RTS/CTS exchange, Clear Channel Assessment (CCS), and promiscuous channel monitoring. Finally, the COLA algorithm has a short critical path of just 10 instructions, and it is free of heuristic parameters, which will facilitate practical use.

State minimization for concurrent system analysis based on state space exploration

A fundamental issue in the automated analysis of concurrent systems is the efficient generation of the reachable state space. Since it is not possible to explore all the reachable states of a system if the number of states is very large or infinite, we need to develop techniques for minimizing the state space. This paper presents our approach to cluster subsets of states into equivalent classes. We assume that concurrent systems are specified as communicating state machines with arbitrary data space. We describe a procedure for constructing a minimal reachability state graph from communicating state machines. As an illustration of our approach, we analyze a producer-consumer program written in Ada.<<ETX>>

An efficient state space generation for the analysis of real-time systems

State explosion is a well-known problem that impedes analysis and testing based on state-space exploration. This problem is particularly serious in real time systems because unbounded time values cause the state space to be infinite even for simple systems. The author presents an algorithm that produces a compact representation of the reachable state space of a real time system. The algorithm yields a small state space, but still retains enough information for analysis. To avoid the state explosion which can be caused by simply adding time values to states, our algorithm uses history equivalence and transition bisimulation to collapse states into equivalent classes. Through history equivalence, states are merged into an equivalence class with the same untimed executions up to the states. Using transition bisimulation, the states that have the same future behaviors are further collapsed. The resultant state space is finite and can be used to analyze real time properties. To show the effectiveness of our algorithm, we have implemented the algorithm and have analyzed several example applications.