Sumit Goswami

Learning Age and Gender of Blogger from Stylistic Variation

We report results of stylistic differences in blogging for gender and age group variation. The results are based on two mutually independent features. The first feature is the use of slang words which is a new concept proposed by us for Stylistic study of bloggers. For the second feature, we have analyzed the variation in average length of sentences across various age groups and gender. These features are augmented with previous study results reported in literature for stylistic analysis. The combined feature list enhances the accuracy by a remarkable extent in predicting age and gender. These machine learning experiments were done on two separate demographically tagged blog corpus. Gender determination is more accurate than age group detection over the data spread across all ages but the accuracy of age prediction increases if we sample data with remarkable age difference.

Design and implementation analysis of a public key infrastructure-enabled security framework for ZigBee sensor networks

SUMMARY ZigBee is a wireless network technology suitable for applications requiring lower bandwidth, low energy consumption and small packet size. Security has been one of the challenges in ZigBee networks. Public Key Infrastructure (PKI) provides a binding of entities with public keys through a Certifying Authority (CA). Public key cryptography using public–private key pairs can be used for ensuring secure transmission in a network. But large size of public and private keys and memory limitations in ZigBee devices pose a problem for using PKI to secure communication in ZigBee networks. In this paper, we propose a PKI enabled secure communication schema for ZigBee networks. Limited memory and power constraints of end devices restrict them from storing public keys of all other devices in the network. Large keys cannot be communicated due to limited power of the nodes and small transmission packet size. The proposed schema addresses these limitations. We propose two algorithms for sending and receiving the messages. The protocols for intercommunication between the network entities are also presented. Minor changes have been introduced in the capabilities of devices used in the ZigBee networks to suit our proposed scheme. Network adaptations depending on different scenarios are discussed. The approach adopted in this paper is to alter the communication flow so as to necessitate minimum memory and computational requirements at the resource starved end points. In the proposed PKI implementation, end devices store the public keys of only the coordinator which in turn holds public keys of all devices in the network. All communication in our scheme is through the coordinator, which in the event of failure is re-elected through an election mechanism. The performance of the proposed scheme was evaluated using a protocol analyzer in home automation and messenger applications. Results indicate that depending on the type of application, only a marginal increase in latency of 2 to 5 ms is introduced for the added security. Layer wise traffic and packets captured between devices were analyzed. Channel utilization, message length distribution and message types were also evaluated. The proposed protocols performance was found to be satisfactory on the two tested applications. Copyright

A fuzzy based approach to stylometric analysis of blogger's age and gender

Fuzzy logic deals with partial truth. A fuzzy based approach to blog analysis, on the basis of various feature words, allows us to determine the degree to which a bloggers style belongs to a particular age or gender group. Each blog was represented by a set of normalized word frequencies of selected feature words in it. Using membership values obtained from applying Fuzzy C-Means (FCM) algorithm to these blog representations, we can call the bloggers style to belong weakly, fairly, strongly or very strongly to a particular class. The advantage of using fuzzy logic for this problem is that a weak belonging to a particular class means that there is a decent belonging to the other class (es). Hence when a search or query is carried out, no useful blog will be left out of the results for that other class (es).

Efficient detection of public key infrastructure-based revoked keys in mobile ad hoc networks

Key revocation involves secure and efficient managing of the information about compromised keys. Spreading the information of revoked keys to the receivers of the key is a challenging task in public key infrastructure (PKI). PKI is more suitable for wired Internet infrastructure and lacks any tailor-made protocols for extension over an ad hoc network. The paper presents a Mobile Ad hoc Key Revocation Server (MAKeRS) scheme which proposes to improve the performance and reliability of the system. Simulation shows that the concept presented in the paper is more reliable, faster, and scalable than the existing usage of PKI over Ad hoc networks. It proposes auto-creation of zone of network availability (ZoNA) by each MAKeRS, which holds the revocation list and is the best service provider in its zone. A node automatically updates the identity of the key revocation server when it enters a new ZoNA. Each node maintains a list of identities of the key revocation servers sorted in order of their communication overhead. This list is regularly updated based on the broadcast from the servers and also gets modified based on the mobility of nodes and servers. The various scenarios of mobility of nodes and servers are considered and the scheme is designed to suit such scenarios in an optimum way. It reduces the time to gain information about the revocation list and ensures availability and, thus, improvement of the system as a whole. Hence, the proposed system results in scalable, reliable, and faster PKI infrastructure and will be attractive for the mobile Ad hoc network (MANET) users who frequently connect to the Internet for secured transactions. We discuss the architecture as well as the performance of our scheme compared to the popular existing scheme. However, our scheme does not call for the entire change in PKI, but is compatible with the existing scheme. Our simulations show that the proposed scheme is better for key revocation. Copyright

Geographic server distribution model for key revocation

Key management is one of the important issues in ensuring the security of network services. The aim of key management is to ensure availability of the keys at both the receiver’s and the sender’s ends. Key management involves two aspects: key distribution and key revocation. Key distribution involves the distribution of keys to various nodes with secrecy to provide authenticity and privacy. Key revocation involves securely and efficiently managing the information about the keys which have been compromised. This paper presents the geographic server distributed model for key revocation which concerns about the security and performance of the system. The concept presented in this paper is more reliable, faster and scalable than the existing Public Key Infrastructure (PKI) framework in various countries, as it provides optimization of key authentication in a network. It proposes auto-seeking of a geographically distributed certifying authority’s key revocation server, which holds the revocation lists by the client, based on the best service availability. The network is divided itself into the strongest availability zones (SAZ), which automatically allows the new receiver to update the address of the authentication server and replace the old address with the new address of the SAZ, in case it moves to another location in the zone, or in case the server becomes unavailable in the same zone. In this way, it reduces the time to gain information about the revocation list and ensures availability and, thus, improvement of the system as a whole. Hence, the proposed system results in scalable, reliable and faster PKI infrastructure and will be attractive for the users who frequently change their location in the network. Our scheme eases out the revocation mechanism and enables key revocation in the legacy systems. It discusses the architecture as well as the performance of our scheme as compared to the existing scheme. However, our scheme does not call for the entire change in PKI, but is compatible with the existing scheme. Our simulations show that the proposed scheme is better for key revocation.

Dividing PKI in strongest availability zones

Key management involves two aspects: key distribution and key revocation. This paper presents the geographic server distributed model for key revocation which concerns about the security and performance of the system. The concept presented in this paper is more reliable, faster and scalable than the existing revocation techniques used in Public Key Infrastructure (PKI) framework in various countries, as it optimises key authentication in a network. It proposes auto-seeking of a geographically distributed certifying authoritys key revocation server, which holds the revocation lists by the client, based on the best service availability. The network is divided itself into the strongest availability zones (SAZ), which automatically allows the new receiver to update the address of the authentication server and replace the old address with the new address of the SAZ, in case it moves to another location in the zone, or in case the server becomes unavailable in the same zone. Our scheme eases out the revocation mechanism and enables key revocation in the legacy systems.

Cognitive prediction of end-to-end bandwidth utilisation in a non-QoS video conference

The most sought after bandwidth killer application on networks has been video conference. For a complex network, specifically based within an organization, implementing quality of service (QoS) is administratively not always feasible as the priorities are regulated. Predicting future network traffic in a non-QoS implemented network by using information about the source, destination and application can give preparatory time to make the network ready for unstable and random demands. The paper uses machine learning techniques to predict the bandwidth utilization of an end-to-end video conference session. Experimental results in this paper show that these features work well in detecting the bandwidth utilization. These experiments were done on a corpus of 24,000 video conference connections. The cognition is based on experimenting on features such as time of call, source, destination, call type, expected duration and cause codes. The result is based on combination of all these features which gave an accuracy of more than 78% on real traffic using two of the common classifiers - k-nearest neighbors and tree based classifier. Support vector machine (SVM) and Naive Bayes gave lower learning accuracy. Prediction results were also obtained by varying the combination of features to detect the predominating features in the cognition. It has been established that the bandwidth at which the connection is established is not entirely dependent on the source and destination but the other features also play a role in deciding the bandwidth of the connection. The prediction accuracy further increases if video calls are allowed only at discrete pre-designated bandwidth levels.

A PKI Adapted Model for Secure Information Dissemination in Industrial Control and Automation 6LoWPANs

Wireless sensor nodes have a wide span of applications ranging from industrial monitoring to military operations. These nodes are highly constrained in terms of battery life, processing capabilities, and in-built memory. Industrial wireless sensor networks (IWSNs) have to meet the constraints and peculiarities of industrial environments to ensure synchronization with parallel production processes. Applications of WSNs in industrial communication vary from condition monitoring and sensing to process automation. The 6LoWPAN standard enables efficient utilization of IPv6 protocol over low-power wireless personal area networks (LoWPANs). The use of 6LoWPANs for industrial communication necessitates the fulfillment of special QoS and security. We examine the aspect of secured information dissemination for industrial control and automation processes in this paper. Researchers have proposed several schemes to secure transfer of data over the Internet. Public key infrastructure (PKI) is one of the most popular security schemes being used in the present scenario. The hostile deployment scenarios of 6LoWPANs and resource constraints of the nodes necessitate the presence of a robust security mechanism to safeguard the communication. In this paper, we propose an integration scheme for PKI and 6LoWPAN to meet the enhanced security needs of industrial communication. The approach is to delegate a major portion of key management activity to the edge routers (gateway) of the LoWPAN and limit the involvement of the end nodes to minimal communication with the edge router. We do not propose a change in the current PKI, but we put forth a scheme to facilitate the integration of PKI to 6LoWPAN in an efficient manner. The effectiveness of the proposed algorithm was evaluated using a protocol analyzer for normal 6LoWPAN traffic as well as HUI HC-01 compressed traffic. A marginal increase of 2% in channel utilization was observed, which scaled down to 1% using HUI HC-01 compression. The results indicated that the proposed algorithm can be implemented for industrial control and automation networks without any speed, security, or performance tradeoffs.

Cognitive correlation of source-destination pair in a video conference network using call attributes

The paper proposes cognitive learning technique for predicting the destination in a video conference being held over an organizational network. The dataset comprised of 22801 connectivity records of video conferences held during the year 2010-2013. Naive Bayes, k-NN and decision tree were trained on the dataset and the performance of the learning algorithms were evaluated. The destination has been predicted with an accuracy of 58.8% over the entire dataset and with 60.1% accuracy over a subset of the dataset. The results indicated deviation from machine learning trends and some of the reasons for deviations have been analyzed and presented while a few had been left out as research problem. There is scope for application of the presented learning technique in the areas of network anomaly detection, network visualization and connectivity prediction.

Multivariate Data Fusion-Based Learning of Video Content and Service Distribution for Cyber Physical Social Systems

Integration of physical processes with the computing world is driving newer challenges for networking frameworks. Cyber physical social systems (CPSSs) are another upcoming paradigm that encompasses the ever-growing interaction between the physical, social, and cyber worlds. As communication networks form the basis of these interactions, a cognitive evaluation of networks is called for. This CPSS driven network evolution was a direction motivating this paper. With the implementation of the next generation networks, traffic from real-time interactive services, such as video conferencing, is surpassing those of conventional transactional services. As such multimedia data transportation over IP networks has stringent quality constraints in terms of required bandwidth, latency, and jitter, legacy networks with no quality of service face challenges in terms of performance. We attempt to perform a multivariate analysis of video call record data collected from a wide area organizational network over a period of time. Learning-based prediction is attempted by training four classifiers: naïve Bayes,