Dr Uday Tupakula

The Internet of Things (IoT) is increasingly being used in applications ranging from precision agriculture to critical national infrastructure by deploying a large number of resource-constrained devices in hostile environments. These devices are being exploited to launch attacks in cyber systems. As a result, security has become a significant concern in the design of IoT-based applications. In this article, we present a security architecture for IoT networks by leveraging the underlying features supported by software-defined networks (SDNs). Our security architecture not only restricts network access to authenticated IoT devices but also enforces fine granular policies to secure the flows in the IoT network infrastructure. The authentication is achieved using a lightweight protocol to authenticate IoT devices. Authorization is achieved using a dynamic policy driven approach. Such an integrated security approach involving authentication of IoT devices and enables authorized flows to protect IoT networks from malicious IoT devices and attacks. We have implemented and validated our architecture using ONOS SDN Controller and Raspbian Virtual Machines, and demonstrated how the proposed security mechanisms can counteract malware packet injection, DDoS attacks using Mirai, spoofing/masquerading, and man-in-the-middle attacks. An analysis of the security and performance of the proposed security mechanisms and their applications is presented in this article.

Earlier, we have proposed an automated model to minimise DDoS attacks in single ISP domain and extended the model to multiple ISP domains. Our approach has several advanced features to minimise DDoS attacks in the internet. The focus of this paper is twofold: firstly, to present a detailed description of the design and implementation of the proposed model and second to discuss and analyse the extensive set of results obtained from the implementation and simulations. We describe the prototype implementation of our automated model using NetProwler network intrusion detection system and HP OpenView Network Node Manager. We will also discuss the performance analysis of our model on a large scale using NS2 tool. Both prototype and simulation test results confirm that our approach offers a promising solution against DDoS problem in the internet and the model can be implemented in real time with minor modifications to the existing tools. Copyright © 2007 Inderscience Enterprises Ltd.

We propose a Controller-Agent model that would greatly minimize distributed denial-of-servicfe (DDoS) attacks on the Internet. We introduce a new packet marking technique and agent design that enables us to identify the approximate source of attack (nearest router) with a single packet even in the case of attacks with spoofed source addresses. Our model is invoked only during attack times, and is able to process the victims traffic separately without disturbing other traffic, it is also able to establish different attack signatures for different attacking sources and can prevent the attack traffic at the nearest router to the attacking system. It is simple in its implementation, it has fast response for any changes in attack traffic pattern, and can be incrementally deployed. Hence we believe that the model proposed in this paper seems to be a promising approach to prevent distributed denial-of-service attacks.

The botnet is a group of hijacked computers, which are employed under command and control mechanism administered by a botmaster. Botnet evolved from IRC based centralized botnet to employing common protocols such as HTTP with decentralized architectures and then peer-to-peer designs. As Botnets have become more sophisticated, the need for advanced techniques and research against botnets has grown. In this paper, we propose techniques to detect botnets by analysing network traffic flows. We developed templates for capturing traffic flows with more relevant attributes for botnet detection. Also we make use of the IPFIX standard for the specification of the templates. Hence our techniques can be used to detect different bot families with lesser overheads and are vendor neutral.

Web based applications are very common nowadays where almost every software can be accessible through a web browser in one form or the other. This paper proposes techniques to detect diffierent threats related to web applications by using a hypervisorbased security architecture. The proposed architecture leverages the hypervisor's visibility of the virtual machines' runtime state and traffic ows for securing the web application. The unique feature of the proposed architecture is that it is capable of doing fine granular detection of web application attacks, i.e. to the specific web page level, and protecting the application against zero-day attacks. © 2015, Australian Computer Society, Inc.

There are several challenges for monitoring the patients with specific requirements such as people with dementia. For example, vascular dementia which is caused generally after stroke could result in serious conditions and change of behaviour such as wandering, loss of vision and speech. Although the nursing staff make sincere effort for taking care and monitoring of the patients, it is rare that a nursing staff is allocated to each patient. Hence even a minor lack of attention can lead to havoc situation if any of the patient is found to be missing. This results in high stress for the nursing staff and the hospital management. The aim of this work is to develop techniques for secure monitoring of dementia patients in hospital environments. Our model tracks the patients in real time and can generate alarms if the location of the patients is found to be suspicious. Furthermore, our model makes use of the existing infrastructures to minimize the cost of deployment. Copyright 2014 ACM.

Domain Name System (DNS) is one of the critical services in the current Internet infrastructure. However DNS is vulnerable to a range of attacks. One of the fundamental weaknesses with the existing DNS protocols is that the request and response messages are transmitted on the network as plain text. This paper addresses important threats related to Doman Name System (DNS) using a hypervisor based security architecture. The proposed architecture leverages the hypervisor visibility of the virtual machines' traffic flows to monitor and utilise Virtual Machine Introspection (VMI) techniques to inspect and restore data. It also uses inbuilt snapshot/restore capabilities of the hypervisor to completely restore virtual machines if required. Objective of the proposed architecture is not to actively prevent attacks, but provide a means of identifying different attacks by passively monitoring DNS related conversations coming in and out of virtualised system hosting the DNS. Our model can alert the external monitoring agent(s) or security administrator and actively restore the system if the attack has already compromised the DNS. © 2014, Australian Computer Society, Inc.

Today, cloud computing is one of the popular technologies. In addition to this, most of the hardware that is being shipped today is equipped with the TPM which can be used for realization of trusted platforms. Recently several TPM attestation techniques such as binary attestation and property based attestation techniques have been proposed but there are some fundamental issues that need to be addressed for using these techniques in practice. In this paper we consider an architecture where different services are hosted on the cloud infrastructure by multiple cloud customers (tenants). Then we consider an attacker model that is specific to the cloud and some of the challenges with the current TPM based attestation techniques. We will also propose a novel trust enhanced security model for cloud which overcomes the challenges with the current TPM based attestation techniques and efficiently deals with the attacks in the cloud. In our model, the cloud service provider is used as the Certification Authority (CA) for the tenant virtual machines. The CA only certifies the basic security properties which are the assurance on the traffic originating from the tenant virtual machine and validation of the tenant virtual machine transactions. The components of the CA monitor the interactions of the tenant virtual machine for the certified properties. Since the tenant virtual machines are running on the cloud service provider infrastructure, it is aware of the dynamic changes to the tenant virtual machine. The CA can terminate the ongoing transactions and/or dynamically isolate the tenant virtual machine if there is a variation in the behaviour of the tenant virtual machine from the certified properties. Hence our model can be used to address the challenges with the current TPM based attestation techniques and efficiently deal with the attacks in the cloud. We will present implementation of our model on Xen and how it deals with the attacks in different attack case scenarios. We will also show that our model is beneficial for the cloud service providers, tenants and tenant customers. © 2012 IEEE.

Significant developments in the recent times have led to an increasing use of mobile devices such as smart phones in accessing Internet services and applications over wireless networks. In this paper, we propose a security architecture for counteracting denial of service attacks in Beyond 3G (B3G) network architecture with mobile nodes. We describe the system architecture and discuss the different cases of attack scenarios involving the mobility of the attacking and victim nodes. Our proposed solution takes into account practical issues such as limited resources of the mobile nodes. It has distinct advantages such as monitoring of the traffic to the victim node and the attack traffic being dropped before reaching the victim; the ability to traceback the attacking node and prevent the attack at the home agent or foreign agent that is closer to the attacking node; and the ability to deal with dynamic changes in attack traffic patterns. We also present an analysis of our proposed architecture as well as simulation results. © 2011 IEEE.

Virtualisation is one of the important technologies for the realisation of cloud computing. A Virtual Machine Monitor (VMM) is an additional software layer which has complete control on the physical resources and enables to run multiple operating systems on a scalable computer. Recently some of the techniques have been proposed to develop Trusted Virtual domains. A trusted virtual domain (TVD) enables grouping of related virtual machines running on separate physical machine into a single network domain with a unified security policy. In this paper we analyze the security issues related to TVD and propose security techniques to deal with the attacks in TVD. © 2011 IEEE.

Today, cloud computing is one of the increasingly popular technology where the customer can use the resources of the cloud services providers to perform their tasks and only pay for the resources they use. The customer virtual machines in the cloud are vulnerable to different types of attacks. In this paper we propose techniques for securing customer virtual machines from different types of attacks in the Infrastructure as a Service cloud and describe how this can be achieved in practice. Our model enables to differentiate attack traffic originating from each virtual machine even if multiple virtual machines on a VMM are sharing a single IP address. © 2011 IEEE.

In this paper we propose comprehensive security architecture called VICTOR to deal with different types of attacks on virtual machines. Our model takes into account the specific characteristics of operating system and applications running in each virtual machine (VM) at a fine granular level to deal with the attacks. Our architecture has several components such as entity validation, intrusion detection engine and dynamic analyzer. The entity validation component is used in the detection of attack traffic with spoofed source address, secure logging, and capturing information of the operating system and applications running in the virtual machines. The intrusion detection engine component is used for detection of known attacks and suspicious behaviour by monitoring the incoming and outgoing traffic of virtual machines. The dynamic analyzer is used for detection and validation of suspicious processes, detection of zero day attacks and fine granular isolation of malicious process or application that is generating the attack traffic. © 2011 IEEE.

The security protocols for WLAN such as WEP have fundamental weakness which can be exploited by the attacker to obtain unauthorized access to the wireless networks and generate attacks. In this paper, we propose a security architecture for counteracting denial of service attacks in wireless based network architecture with mobile nodes. We describe the system model and discuss the different cases of attack scenarios involving the mobility of the attacking and victim nodes. We describe how mobile IP protocol in conjunction with our model can be used to deal efficiently with the attacks on mobile nodes. © 2011 ACM.

A trusted virtual domain (TVD) enables grouping of related virtual machines running on separate physical machine into a single network domain with a unified security policy. Since the virtual machines can be running different operating systems and applications, the attacker can generate attacks in the TVD by exploiting a single vulnerability in any of the operating systems or applications. Our aim in this paper is to consider the design choices and develop an intrusion detection architecture that would enable efficient detection and prevention of different types of attacks in such a TVD based distributed environments. The proposed architecture can capture the knowledge of the operating systems and applications at fine granular level and isolate the malicious entities that are generating the attack traffic. Our model takes into account the security policies that are specific to the virtual machine as well as security policies of the trusted virtual domains to deal with the attacks efficiently. © 2010 IEEE.

Denial of service (DoS) attacks are one of the complex problems in the current Internet. In this paper, we propose a system, DoSTRACK, that can efficiently deal with the TCP SYN and reflection Distributed Denial of Service (DDoS) attacks. We also describe a prototype implementation of our model with HP OpenView Network Node Manager (NNM) and discuss how our model can be beneficial to the DDoS victim and the ISP. Copyright 2009 ACM.

In last five years, several trust models have been proposed to enhance the security of Mobile Ad hoc Networks (MANET). Nevertheless, these trust models fail to express the notion of ignorance during the establishment of trust relationships between mobile nodes. Furthermore, they lack a well-defined approach to defend against the issues resulting from recommendations. In this paper, we propose a novel subjective logic based trust model that enables mobile nodes to explicitly represent and manage ignorance as uncertainty during the establishment of trust relationships with other nodes. Our model defines additional operators to subjective logic in order to address the ignorance introduced between mobile nodes (which have already established trust relationships) as a result of mobility-induced separation. Second, we demonstrate on how mobile nodes formulate their opinions for other nodes based on the evidence collected from the benign and malicious behaviors of those nodes. We then describe on how mobile nodes establish trust relationships with other nodes using the opinions held for those nodes. Depending on the policies defined, these relationships are then used by our model to enhance the security of mobile communications. Third, we propose a novel approach to communicate recommendations by which no explicit packets or additional headers are disseminated as recommendations. This allows our model to defend against recommendation related issues such as free-riding, honest-elicitation, and recommender's bias. Finally, we demonstrate the performance of our model through NS2 simulations. Copyright © 2008 ACM.

Recently several techniques that provide different degree of anonymity have been proposed for wired and wireless communication. Although, the recently proposed techniques are successful in achieving high degree of anonymity, there are some disadvantages associated with the proposed techniques. In this paper we analyze the flooding and packet drop attacks in mobile ad hoc networks that support anonymous communication. Then we propose a novel technique to deal with the flooding attacks. Our approach can efficiently identify and isolate the malicious node that floods the network. In addition, our technique provides a mechanism to identify the benign behavior of an expelled node and rejoins the expelled node back into the network. Furthermore, our approach does not require any additional packets to communicate the behavior of the flooding node and hence does not incur any additional overhead. Finally we validate the performance analysis of our technique through NS2 simulations. © 2007 IEEE.

In recent years, several secure routing protocols have been proposed to secure communications among nodes in mobile ad hoc networks. However, they are not tailored to defend against Denial of Service (DoS) attacks such as flooding and packet drop attacks. This has led to the development of models that target cooperation among nodes. These models either fail to protect against flooding attacks or only defend against greedy nodes that drop packets to save battery resources. The main shortcoming of cooperation models is that they fail to evaluate the trustworthiness for other nodes. In this paper, we propose a Trust Integrated Cooperation Architecture which consists of an obligation-based cooperation model known as fellowship to defend against both flooding and packet drop attacks. In our architecture, fellowship enhances its security decisions through a trust model known as Secure MANET Routing with Trust Intrigue (SMRTI). In comparison with related models, SMRTI deploys a novel approach to communicate recommendations such that the deployed approach is free from well-known issues such as honest elicitation, free riding, bias of a recommender, and additional overhead. © 2007 IEEE.

Security is paramount in Mobile Ad-hoc Networks (MANET) as they are not conducive to centralized trusted authorities. Several solutions have been proposed for MANET in the areas of key management, secure routing, nodal cooperation, and trust management. Nevertheless, MANET lacks a unified architecture to take advantage of the deployed security models. In this paper, we propose Trust Enhanced security Architecture for MANET (TEAM), in which a trust model is overlaid on the following security models - key management mechanism, secure routing protocol, and cooperation model. We briefly present the operation of our architecture and then we detail the system operation of our novel trust and cooperation model, which we call as Secure MANET Routing with Trust Intrigue (SMRTI) and fellowship respectively. SMRTI captures the evidence of trustworthiness for other nodes from the security models, and in return assists them to make better security decisions. Unlike related trust models, SMRTI captures recommendations in such a way that it eliminates both freeriding and honest-elicitation problems. In comparison with related cooperation models, fellowship model defends against both flooding and packet drop attacks. It can efficiently identify and isolate both malicious and selfish nodes that fail to share the communication channel or forward packets for other nodes. Furthermore, our models do not rely on any centralized authority or tamper-proof hardware. Simulation results confirm that our models enhance the performance of TEAM. © 2007 IEEE.

Recently several trust and reputation models have been proposed to enhance the security of mobile ad hoc networks. In these models, recommendations are circulated by forwarding explicit messages or introducing extra message headers. Apart from incurring additional overhead, the recommendations are prone to issues such as recommender's bias, honest-elicitation, and free-riding. In this paper, we propose a trust model to enhance the security of mobile ad hoc networks and to address the issues related to recommendations. The model uses only trusted routes for communication, and isolates malicious nodes depending on the evidence collected from direct interactions and recommendations. It deploys a novel approach for communicating recommendations such that they are free from recommender's bias, honest-elicitation, and free-riding. Simulation results confirm the effectiveness of our model. © 2007 IEEE.

Today's Internet is extremely vulnerable to Distributed Denial of service (DDoS) attacks. There is tremendous pressure on the sites performing online business and ISP's to protect their networks from DDoS attacks. Recently, several novel traceback techniques have been proposed to trace the approximate spoofed source of attack. Each proposed traceback technique has some unique advantages and disadvantages over the others. In this paper we will consider some of the novel traceback techniques and focus our discussion i) to raise some of the real time issues that can be addressed in the further research and ii) from the attackers perspective on how to generate DDoS attacks and remain untraced even if any of the traceback technique is deployed in the Internet. We will also demonstrate how attacks can be further amplified if ICMP traceback technique is deployed in the Internet and discuss techniques to minimise the additional attack traffic. We believe that the networks tend to become complex and more vulnerable to DDoS attacks if some of the proposed traceback techniques are deployed in the Internet. © 2006, Australian Computer Society, Inc.

In this paper we discuss techniques to prevent Distributed Denial of Service (DDoS) attacks within the ISP domain and extend the scheme to prevent the attack in multiple ISP domains. With a new packet marking technique and agent design, our model is able to identify the approximate source of attack with a single packet and has many features to minimise DDoS attacks. © 2003 by Springer Science+Business Media Dordrecht.