Md. Shohrab Hossain

Stochastic Properties and Application of City Section Mobility Model

Node mobility has a direct impact on the performance evaluation of various network mobility protocols. Unfortunately, most of the analysis on mobility protocols used Random Waypoint mobility model which does not represent real-world movement patterns of mobile nodes. In this paper, we have analyzed City Section mobility model, a realistic mobility model for the movement in the city streets. We have developed an analytical model to derive certain stochastic properties, such as, expected epoch length, expected epoch time, expected number of subnet crossings and subnet residence time of this model. Finally, we have applied the model to calculate the signaling cost of NEMO BSP and compared it with Random Waypoint mobility model. Results show that the use of realistic mobility model leads to better estimation of the signaling cost of network mobility protocol.

On the routing in Flying Ad Hoc Networks

The usage of Unmanned Aerial Vehicles (UAVs) is increasing day by day. In recent years, UAVs are being used in increasing number of civil applications, such as policing, fire-fighting, etc in addition to military applications. Instead of using one large UAV, multiple UAVs are nowadays used for higher coverage area and accuracy. Therefore, networking models are required to allow two or more UAV nodes to communicate directly or via relay node(s). Flying Ad-Hoc Networks (FANETs) are formed which is basically an ad hoc network for UAVs. This is relatively a new technology in network family where requirements vary largely from traditional networking model, such as Mobile Ad-hoc Networks and Vehicular Ad-hoc Networks. In this paper, Flying Ad-Hoc Networks are surveyed along with its challenges compared to traditional ad hoc networks. The existing routing protocols for FANETs are then classified into six major categories which are critically analyzed and compared based on various performance criteria. Our comparative analysis will help network engineers in choosing appropriate routing protocols based on the specific scenario where the FANET will be deployed.

Performance evaluation of multihomed NEMO

Mobile networks can be formed in bus, train, aircrafts, satellites with a wide variety of on-board IP-enabled devices and Network Mobility (NEMO) protocols are required to support uninterrupted services to ongoing sessions. Earlier works have not demonstrated seamless handover for NEMO architecture. In this work, we proposed a handover scheme for NEMO that exploits the multi-homing feature of the Mobile Router and uses make-before-break strategy to ensure seamless handover for NEMO. Using experimental testbed, we have presented a thorough handoff performance evaluation of multihomed NEMO and compared it with basic NEMO. Results demonstrate that the proposed multihomed NEMO outperforms the basic NEMO while achieving seamless handover.

Scheduling internet of things applications in cloud computing

Internet of Things (IoT) is one of the greatest technology revolutions in the history. Due to IoT potential, daily objects will be consciously worked in harmony with optimized performances. However, today, technology is not ready to fully bring its power to our daily life because of huge data analysis requirements in instant time. On the other hand, the powerful data management of cloud computing gives IoT an opportunity to make the revolution in our life. However, the traditional cloud computing server schedulers are not ready to provide services to IoT because IoT consists of a number of heterogeneous devices and applications which are far away from standardization. Therefore, to meet the expectations of users, the traditional cloud computing server schedulers should be improved to efficiently schedule and allocate IoT requests. There are several proposed scheduling algorithms for cloud computing in the literature. However, these scheduling algorithms are limited because of considering neither heterogeneous servers nor dynamic scheduling approach for different priority requests. Our objective is to propose dynamic dedicated server scheduling for heterogeneous and homogeneous systems to efficiently provide desired services by considering priorities of requests. Results show that the proposed scheduling algorithm improves throughput up to 40 % in heterogeneous and homogeneous cloud computing systems for IoT requests. Our proposed scheduling algorithm and related analysis will help cloud service providers build efficient server schedulers which are adaptable to homogeneous and heterogeneous environments by considering system performance metrics, such as drop rate, throughput, and utilization in IoT.

Cost Analysis of Mobility Management Entities of SINEMO

Seamless IP-diversity based NEtwork MObility (SINEMO) was proposed to address a number of drawbacks of the Network Mobility (NEMO) protocol for mobility management of a number of hosts moving together in a Local Area Network. Increasing number of mobile hosts results in higher level of signalling cost on the mobility agents in a mobility protocol. Previous works on cost analysis of mobility protocols have considered the cost on some of the mobility agents, but did not consider all possible costs for mobility management, resulting in incomplete cost estimation. In this paper, we have developed analytical models to estimate costs of mobility management as functions of network size, mobility rate, traffic rate and data volume for all the entities of SINEMO. Numerical results, comparing the cost between mobility entities of SINEMO and NEMO, reveal that SINEMO has higher efficiency and lower cost than NEMO. Our comprehensive cost model can be used as a framework for estimating total cost of key mobility management entities of different handover protocols, and can aid in decision making to choose the most efficient protocol for future all-IP mobile and wireless networks.

Malware detection in Android by network traffic analysis

A common behavior of mobile malware is transferring sensitive information of the cell phone user to malicious remote servers. In this paper, we describe and demonstrate in full detail, a method for detecting malware based on this behavior. For this, we first create an App-URL table that logs all attempts made by all applications to communicate with remote servers. Each entry in this log preserves the application id and the URI that the application contacted. From this log, with the help of a reliable and comprehensive domain blacklist, we can detect rogue applications that communicate with malicious domains. We further propose a behavioral analysis method using syscall tracing. Our work can be integrated with be behavioral analysis to build an intelligent malware detection model.

Multi class traffic analysis of single and multi-band queuing system

To facilitate higher bandwidth for multimedia traffic, modern routers support simultaneous multi-band communication, leading to less interference, higher capacity and better reliability. However, there is lack of quantitative evaluation to judge whether multi-band is better than single band router in realistic scenarios. Our objective is to propose a scheduling algorithm for multi-band routers and compare single band and multi-band system with different allocation policies. We have used different scheduling algorithms for multi-band routers which transmit different classes of traffic through different frequency bands, thereby achieving improved performance. By comparing multi-band and single band mobile router performances, we have found out that one of them is not always better than the other although multi-band is expected to have better performance.

Cost analysis of mobility entities of hierarchical mobile IPv6

IETF proposed Hierarchical Mobile IPv6 (HMIPv6) to support host mobility and mobility management involves signaling costs at various mobility entities of the network. Widespread use of IP-enabled mobile devices have resulted in increase in number of mobile users and the signaling cost on underlying mobility entities have increased significantly, which will result in their performance degradation. However, there has been no comprehensive cost analysis of mobility protocol entities that considers all possible costs. In this paper, we have developed analytical models to estimate total costs of key mobility management entities of HMIPv6. We have presented numerical results to demonstrate the impact of network size, mobility rate, traffic rate and data volume on these costs and the percentage overhead on the mobility entities. Our results show that a significant amount of resources are required by the mobility entities for transmission, processing of various signaling messages, as well as searching location database. Our cost analysis will thus be helpful for military applications in estimating actual resource requirements for on-board IP-enabled devices in military vans, tanks, helicopters that require mobility management especially while in operation.

Behavioral malware detection approaches for Android

Android, the fastest growing mobile operating system released in November 2007, boasts of a staggering 1.4 billion active users. Android users are susceptible to malicious applications that can hack into their personal data due to the lack of careful monitoring of their in-device security. There have been numerous works on devising malware detection methods. However, none of earlier works are conclusive enough for direct application and lack experimental validation. In this paper, we have investigated the natures and identities of malicious applications and devised two novel detection approaches for detection: network-based detection and system call based detection approaches. To evaluate our proposed approaches, we performed experiments on a subset of 1260 malwares, acquired from Android Malware Genome Project, a malware database created by Y. Zhou et al. [1] and 227 non-malware (benign) applications. Results show that our system call based approach is able to detect malwares with an accuracy of 87% which is quite significant in general malware detection context. Our proposed detection approaches along with the experimental results will provide security professionals with more precise and quantitative approaches in their investigations of mobile malwares on Android systems.

A cost analysis framework for claimer reporter witness based clone detection schemes in WSNs

A number of Claimer-Reporter-Witness based schemes (also called witness node based schemes) have been proposed to solve the problem of clone node attack or node replication attack which brings severe threats to WSNs. Node replication attack is an application independent attack in which an adversary first physically captures and then compromises a sensor node in the network and after creating clones or replicas of the compromised nodes; he/she surreptitiously deploys them at strategic positions of the network. The approaches used by the existing Claimer Reporter Witness-based detection schemes tradeoff between the higher detection probability for detecting clones and various overheads. The communication costs along with the memory and processing costs need to be measured to find out the gain from tradeoff. This is the first such work in which we have developed analytical framework to measure various costs for our proposed distributed clone detection protocol RAND and the other existing distributed schemes for clone detection. Moreover, we have also presented numerical results to demonstrate the impact of network size and the number of selected witnesses on these costs. Our results show that a significant amount of resources are required by the chosen CRW based schemes for detecting clones. Our comprehensive cost model can be used as a framework for estimating the costs of different CRW based protocols and can aid the researchers to choose the most efficient protocol for detecting clones in the network as well as estimate the resource requirements of the network to develop effective schemes in future.