I.Next Generation Wireless Communications Systems
- Transmission Protocols
A)TCP Performance in Satellite Communications Systems
B)Handoff Management in Wireless Heterogeneous Networks
- Mobility Management and Quality of Service Routing
A)Traffic Engineering in Satellite Communications Systems
B)Mobility Management in Mobile IPv6
C)Routing in VANET Networks
D)Efficient Use of Energy in Sensor Networks
- On-Demand Multimedia Transmission
II.Internet Security
- Early Detection of Internet Worms
- Detection of Denial of Service Attacks in Hybrid Wired/Wireless Networks
- Intrusion Detection Systems for Encrypted Attacks
Next Generation Wireless Communications Systems
Transmission Protocols
TCP Performance in Satellite Communication Systems
In this research work, we examine some issues that affect theefficiency and fairness of the Transmission Control Protocol(TCP), the backbone of Internet protocol communication, inmulti-hops satellite network systems. We propose a scheme thatallows satellite systems to automatically adapt to any change in the number of active TCP flows due to handover occurrence, thefree buffer size, and the bandwidth-delay product of the network. An extension of this research work to terrestrial networks, particularly to “one-to-many” network topologies where a set of servers provide a potential number of users with a plethora of services (ex. Video-on-Demand -- VoD), is also examined.
The proposed scheme has two major design goals: Increasingthe system efficiency and improving its fairness. The systemefficiency is controlled by matching the aggregate traffic rateto the sum of the link capacity and total buffer size. On theother hand, the system min-max fairness is achieved by allocating bandwidth among individual flows in proportion with their Round Trip Times (RTTs).The proposed scheme is dubbed Recursive, Explicit, and FairWindow Adjustment (REFWA). The extended version of REFWA to terrestrial networks is called Terrestrial- REFWA (T-REFWA).
Simulation results elucidate that the REFWA and T-REFWA schemes substantiallyimprove the system fairness, reduce the number ofpacket drops, and make better utilization of the bottleneck link.The results demonstrate also that the proposed schemes workproperly in more complicated environments where connectionstraverse multiple bottlenecks and the available bandwidth maychange over data transmission time.
Handoff Mangament in Wireless Heterogeneous Networks
Recent trends in telecommunication industry are toward the development of ubiquitousinformation systems where the provision of a plethora of advanced multimedia services should bepossible regardless of time and space limitations. An efficient and seamless delivery ofmultimedia services over various types of wireless networks is still a challenging task. Theunderlying difficulty consists in the disparity in the bandwidth availability over each networktype. Indeed, the fundamental challenge upon a handoff phenomenon in a heterogeneous wirelessnetwork consists in an efficient probing of the bandwidth availability of the new network,followed by a prompt adjustment of the data delivery rate.
In this research work, we design and evaluate a cross layer approach that involves five layers, namely physical,data link, application, network, and transport layers. The three former layers are used to anticipatethe handoff occurrence and to locate the new point of attachment to the network. Based on theirfeedback, the transport layer is used then to probe the resources of the new network usinglow-priority dummy packets. This bandwidth probing operation is called Dummy Segment based Bandwidth Probing (DSBP).Being the most widely used protocols for multimedia delivery, this research work addresses multimedia applications based on TCP and RTP protocols. The design of thewhole cross layer architecture is discussed and enhancements to the two protocols are proposed.
The performance of the enhanced TCP and RTP protocols is evaluated and compared withexisting schemes through extensive simulations. The obtained results are encouraging andpromising for the delivery of multimedia services in heterogeneous wireless networks.
Mobility Management and Quality of Service Routing
Traffic Engineering in Satellite Communications Systems
Due togeographical and/or climatic constraints, the community of futuresatellite users will exhibit a significant variance in its density overthe Globe. This density variance will yield a scenario where somesatellite links are congested while others are underutilized. Toensure an intelligent engineering of traffic over satellite networks,this research work envisions a routing protocol that enables neighboringsatellites to explicitly exchange information on their congestionstatus. A satellite that is about to get congested requests its neighboringsatellites to decrease their data forwarding rates. In response,the neighboring satellites search for less congested paths that donot include the satellite in question and communicate a portionof data, primarily destined to the satellite, via the retrievedpaths. By so doing, congestion, and the resulting packet drops,can be avoided. A better distribution of traffic among satellitescan be guaranteed as well. The proposed scheme is dubbed“Explicit Load Balancing” (ELB) scheme. A set of simulationsis conducted to evaluate the performance of the ELB schemeusing the Network Simulator. In terms of Quality of Service,encouraging results are obtained: better traffic distribution,higher throughput, and lower packet drops. While this research work considers the case of satellite communications systems, the proposed mechanism can be easily applied also to terrestrial networks where nodes are aware of their neighboring nodes.
Mobility Management in Mobile IPv6
Recent trends in telecommunications industry aretoward ubiquitous information technologies. Communicationsover emerging mobile networks are thus gaining a tremendousinterest at both industrial and academic levels.A major concern for mobile networks consists in findingefficient ways to handle the user mobility so that the handoverprocess has minimum effect on user's ongoing sessions. Giventhe dominance of Internet-based applications in next-generationmobile networks, Mobile IP has become an important protocolto accommodate the IP mobility.
To overcome the excessive delay and signaling involved in the first version of Mobile IP, the Hierarchical Mobile IPv6 (HMIPv6)protocol has been introduced. The key concept behind HMIPv6is to locally handle handovers by the usage of an entity calledMobility Anchor Point (MAP). While the new protocol provides amore efficient way for the mobility management in IP networks,it does not control traffic among multiple MAPs in the network.As a result, in many cases the selected MAP is overloaded andextensive delays are experienced during the routing process.
To tackle this problem, this research work portrays a new techniquecalled Dynamic and Efficient MAP Selection (DEMAPS). In theproposed scheme, the most optimum MAP with the lightesttraffic load is selected. This selection is based on an estimationof MAP load transition using the Exponential Moving Average(EMA) method. Simulation results demonstrate that DEMAPScan balance the signaling traffic load efficiently among MAPsand provides a superior network performance compared totraditional HMIP schemes.
Routing in VANET Networks
Internetworking over Vehicle Ad-hoc Networks(VANETs) is getting increasing attention from all major carmanufacturers. The design of effective vehicular communicationsposes a series of technical challenges. Guaranteeing a stable andreliable routing mechanism over VANETs is an important steptowards the realization of effective vehicular communications.
In current ad-hoc routing protocols, the control messages inreactive protocols and route update timers in proactive protocolsare not used to anticipate link breakage. They solely indicatepresence or absence of a route to a given node. Consequently,the route maintenance process at both protocol types is initiatedonly after a link breakage event takes place. This research work arguesthe use of information on vehicle headings to predict a possiblelink breakage event prior to itsoccurrence. Vehicles are groupedaccording to their velocity vectors. When a vehicle shifts toa different group and a route, involving the vehicle, is to bebroken, the proposed protocol searches for a more stable and“more durable” route that includes vehicles from the samegroup. The proposed scheme is dubbed Velocity-Heading basedRouting Protocol (VHRP). The proposed scheme canbe implemented on any existing routing protocol and its performance is evaluated through computer simulations. Simulation resultsindicate that knowledge on the vehicles’ heading adds majorbenefits to routing in terms of reducing the number of linkbreakage events and increasing the end-to-end throughput.
Efficient Use of Energy in Sensor Networks
In this research work, we consider the use of LEO satellites tobuild a global and energy-efficient sensor network. To reducethe total cost of the architecture, a heterogeneous environmentwith two types of nodes is envisioned. Exploiting the multicastcapabilities of the satellites, a set of signaling packets is introduced to guarantee an efficient use of the nodes energy.Nodes are grouped into dynamically changing clusters basedon their distance to cluster heads. Their transmission powersare accordingly adjusted. This directionality-based transmissionfashion helps to largely minimize the energy drainageat each node. At each coverage area, a single cluster headgathers data from its neighboring cluster heads and transmitsit to the satellite on behalf of them. This yields to an efficientuse of the cluster heads energy. From these characteristics, theproposed approach is expected to achieve some savingsin the scarce energy of nodes and thus to increase the lifetimeof the sensor network. By substituting LEO systems with High-Altitude Platforms(HAP) or Unmanned Airborne Vehicles (UAV), a more cost-effective system can be realized.
On-Demand Multimedia Transmission
Since the number of Internet users is rapidly increasing day by day and even the most powerful server system willalways be resource limited, one of the challenges faced by multimedia-on-demand system designers is how to configure asystem that can support a potentially large number of customers and a large multimedia library to satisfy users’ needs ataffordable rates.
In this research work, we propose an approach to provide a significantly scalable multimedia-on-demand service in a multicastenvironment. The basic idea is to repeatedly transmit popular video items on staggered channels. If a request comes inbetween staggered start times, the user joins to the most recently started multicast session and then requests the missingpart from a nearby neighbor. Users must have enough buffer space to buffer data between staggered transmissions.We refer to the proposed architecture as Neighbors-Buffering Based Video-on-Demand (NBB-VoD) architecture.
Based ona combination ofsatellite systems and terrestrial networks, we build also a global, large-scale, and efficient Video-on-Demand (VoD) architecture. A hybrid network made of fixed and mobile nodes is considered. The key idea of the architecture is to service fixednodes according to the NBB policy, while mobile nodes areserved directly from the local server. To allow users to receive their multimedia applications with higher degree of mobility,issues related to mobility management are discussed and a simple scheme is proposed to guarantee a smooth streamingof video data.The entirearchitecture is referred to as a “Theatre in the Sky”.
The importance of the proposed architecture is verified by numerical results. In case of requests coming from fixednodes within the reach of terrestrial networks, analytical results elucidate the good performance of the architecture interms of both increasing the system capacity and reducing the disk-bandwidth requirements. Conducted simulationsindicate how efficient the proposed system is in smoothening handoffs.
Internet Security
Early Detection of Internet Worms
Internet worms pose a serious threat to the Internet. In this research work, we propose a signature-based Intrusion Detection System (IDS) to detect worms. An anomaly analysis module is embedded in the IDS to detect novel worms. The proposed system detects novel worms and instantly generates their signatures, making it possible to check the spread of any kind of worm –known or unknown ones. We envision a two-layer hierarchical architecture comprising of local security managers, metropolitan security managers, and a global security manager. Local managers collect worm-like or suspicious flows and hand them to metropolitan managers. Metropolitan managers then use cluster analysis to sort worms from the suspicious flows. These sorted worms are used to generate worm signatures. The global manager relays the signature to the local managers. We also present a scheme to automatically decide optimum values for the system parameters that best suit the prevailing network state. We evaluate the proposed scheme using real network traffic that contains traces of worms. Through experiments using real network traffic, we show that the proposed method is capable of detecting worms at the early phase of their propagations.
Detection of Denial of Service Attacks in Hybrid Wired/Wireless Networks
Protection of Mobile IP networks from Denial-of-Service (DoS) attacks, a serious security threat in today’s Internet, is a one major step toward making this paradigm a reality. The paper proposes a method to detect DoS attacks, issued from mobile users, in the vicinity of flooding sources and in early stages before they cripple the targeted system. The fundamental challenge in attack detection consists in distinguishing between simple flash events and DoS attacks so as not to deprive innocent users from having legitimate accesses.
In this research work, this distinction is based on the fact that legitimate TCP flows obey the congestion control protocol, whereas misbehaving sources remain unresponsive. Suspicious flows are sent a test feedback and are required to decrease their sending rates. Legitimacy of such flows is decided based on their responsiveness. The scheme performance is evaluated through a set of simulations and encouraging results are obtained: short detection latency and high detection accuracy.
Intrusion Detecion Systems for Encrypted Attacks
Most modern servers run equipped with cryptographicor encrypted protocols such as Secure SocketLayer (SSL),Transport Layer Security (TLS), and Secure Socket protocol(SSH). These encrypted protocols, however, themselvesare subject to exploits, unauthorized access and attacks. Bymonitoring such activities in the network, it is possibleto cut off such attacks before they cause serious damage tothe intended target.The objective of this researchwork is to offer an anomaly based detection system against attackson cryptographic protocols. Upon detecting such an attack,the proposed system will take action against the attacker bytracing back to the attack source. The detection, prevention andtracing back are carried out at distributed monitor stubs whichare simply network-sniffers without modifying incoming oroutgoing network packets.