-
Model Predictive Simulation Using Structured Graphical Models and Transformers
Authors:
Xinghua Lou,
Meet Dave,
Shrinu Kushagra,
Miguel Lazaro-Gredilla,
Kevin Murphy
Abstract:
We propose an approach to simulating trajectories of multiple interacting agents (road users) based on transformers and probabilistic graphical models (PGMs), and apply it to the Waymo SimAgents challenge. The transformer baseline is based on the MTR model, which predicts multiple future trajectories conditioned on the past trajectories and static road layout features. We then improve upon these g…
▽ More
We propose an approach to simulating trajectories of multiple interacting agents (road users) based on transformers and probabilistic graphical models (PGMs), and apply it to the Waymo SimAgents challenge. The transformer baseline is based on the MTR model, which predicts multiple future trajectories conditioned on the past trajectories and static road layout features. We then improve upon these generated trajectories using a PGM, which contains factors which encode prior knowledge, such as a preference for smooth trajectories, and avoidance of collisions with static obstacles and other moving agents. We perform (approximate) MAP inference in this PGM using the Gauss-Newton method. Finally we sample $K=32$ trajectories for each of the $N \sim 100$ agents for the next $T=8 Δ$ time steps, where $Δ=10$ is the sampling rate per second. Following the Model Predictive Control (MPC) paradigm, we only return the first element of our forecasted trajectories at each step, and then we replan, so that the simulation can constantly adapt to its changing environment. We therefore call our approach "Model Predictive Simulation" or MPS. We show that MPS improves upon the MTR baseline, especially in safety critical metrics such as collision rate. Furthermore, our approach is compatible with any underlying forecasting model, and does not require extra training, so we believe it is a valuable contribution to the community.
△ Less
Submitted 27 June, 2024;
originally announced June 2024.
-
Fast exploration and learning of latent graphs with aliased observations
Authors:
Miguel Lazaro-Gredilla,
Ishan Deshpande,
Sivaramakrishnan Swaminathan,
Meet Dave,
Dileep George
Abstract:
We consider the problem of recovering a latent graph where the observations at each node are \emph{aliased}, and transitions are stochastic. Observations are gathered by an agent traversing the graph. Aliasing means that multiple nodes emit the same observation, so the agent can not know in which node it is located. The agent needs to uncover the hidden topology as accurately as possible and in as…
▽ More
We consider the problem of recovering a latent graph where the observations at each node are \emph{aliased}, and transitions are stochastic. Observations are gathered by an agent traversing the graph. Aliasing means that multiple nodes emit the same observation, so the agent can not know in which node it is located. The agent needs to uncover the hidden topology as accurately as possible and in as few steps as possible. This is equivalent to efficient recovery of the transition probabilities of a partially observable Markov decision process (POMDP) in which the observation probabilities are known. An algorithm for efficiently exploring (and ultimately recovering) the latent graph is provided. Our approach is exponentially faster than naive exploration in a variety of challenging topologies with aliased observations while remaining competitive with existing baselines in the unaliased regime.
△ Less
Submitted 25 September, 2023; v1 submitted 13 March, 2023;
originally announced March 2023.
-
Analysis, Detection, and Classification of Android Malware using System Calls
Authors:
Shubham Shakya,
Mayank Dave
Abstract:
With the increasing popularity of Android in the last decade, Android is popular among users as well as attackers. The vast number of android users grabs the attention of attackers on android. Due to the continuous evolution of the variety and attacking techniques of android malware, our detection methods should need an update too. Most of the researcher's works are based on static features, and v…
▽ More
With the increasing popularity of Android in the last decade, Android is popular among users as well as attackers. The vast number of android users grabs the attention of attackers on android. Due to the continuous evolution of the variety and attacking techniques of android malware, our detection methods should need an update too. Most of the researcher's works are based on static features, and very few focus on dynamic features. In this paper, we are filling the literature gap by detecting android malware using System calls. We are running the malicious app in a monitored and controlled environment using an emulator to detect malware. Malicious behavior is activated with some simulated events during its runtime to activate its hostile behavior. Logs collected during the app's runtime are analyzed and fed to different machine learning models for Detection and Family classification of Malware. The result indicates that K-Nearest Neighbor and the Decision Tree gave the highest accuracy in malware detection and Family Classification respectively.
△ Less
Submitted 12 August, 2022;
originally announced August 2022.
-
Improving Privacy and Security in Unmanned Aerial Vehicles Network using Blockchain
Authors:
Hardik Sachdeva,
Shivam Gupta,
Anushka Misra,
Khushbu Chauhan,
Mayank Dave
Abstract:
Unmanned Aerial Vehicles (UAVs), also known as drones, have exploded in every segment present in todays business industry. They have scope in reinventing old businesses, and they are even develo** new opportunities for various brands and franchisors. UAVs are used in the supply chain, maintaining surveillance and serving as mobile hotspots. Although UAVs have potential applications, they bring s…
▽ More
Unmanned Aerial Vehicles (UAVs), also known as drones, have exploded in every segment present in todays business industry. They have scope in reinventing old businesses, and they are even develo** new opportunities for various brands and franchisors. UAVs are used in the supply chain, maintaining surveillance and serving as mobile hotspots. Although UAVs have potential applications, they bring several societal concerns and challenges that need addressing in public safety, privacy, and cyber security. UAVs are prone to various cyber-attacks and vulnerabilities; they can also be hacked and misused by malicious entities resulting in cyber-crime. The adversaries can exploit these vulnerabilities, leading to data loss, property, and destruction of life. One can partially detect the attacks like false information dissemination, jamming, gray hole, blackhole, and GPS spoofing by monitoring the UAV behavior, but it may not resolve privacy issues. This paper presents secure communication between UAVs using blockchain technology. Our approach involves building smart contracts and making a secure and reliable UAV adhoc network. This network will be resilient to various network attacks and is secure against malicious intrusions.
△ Less
Submitted 27 June, 2022; v1 submitted 16 January, 2022;
originally announced January 2022.
-
Proactive DDoS Attack Mitigation in Cloud-Fog Environment using Moving Target Defense
Authors:
Vaishali Kansal,
Mayank Dave
Abstract:
Distributed Denial of Service (DDoS) attacks are serious cyber attacks and mitigating DDoS attacks in cloud is a topic of ongoing research interest which remains a major security challenge. Fog computing is an extension of cloud computing which has been used to secure cloud. Moving Target Defense (MTD) is a newly recognized, proactive security defense that can be used to mitigate DDoS attacks on c…
▽ More
Distributed Denial of Service (DDoS) attacks are serious cyber attacks and mitigating DDoS attacks in cloud is a topic of ongoing research interest which remains a major security challenge. Fog computing is an extension of cloud computing which has been used to secure cloud. Moving Target Defense (MTD) is a newly recognized, proactive security defense that can be used to mitigate DDoS attacks on cloud. MTD intends to make a system dynamic in nature and uncertain by changing attack surface continuously to confuse attackers. In this paper, a novel DDoS mitigation framework is presented to support Cloud-Fog Platform using MTD technique (CFPM). CFPM applies migration MTD technique at fog layer to mitigate DDoS attacks in cloud. It detects attacker among all the legitimate clients proactively at the fog layer and isolate it from innocent clients. CFPM uses an effective request handling procedure for load balancing and attacker isolation procedure which aims to minimize disruption to cloud server as well as serving fog servers. In addition, effectiveness of CFPM is evaluated by analyzing the behavior of the system before and after attack, considering different possible scenarios. This approach is effective as it uses the advantage of both MTD technique and Fog computing paradigm supporting cloud environment.
△ Less
Submitted 3 December, 2020;
originally announced December 2020.
-
Architectural-Space Exploration of Heterogeneous Reliability and Checkpointing Modes for Out-of-Order Superscalar Processors
Authors:
Bharath Srinivas Prabakaran,
Mihika Dave,
Florian Kriebel,
Semeen Rehman,
Muhammad Shafique
Abstract:
Reliability has emerged as a key topic of interest for researchers around the world to detect and/or mitigate the side effects of decreasing transistor sizes, such as soft errors. Traditional solutions, like DMR and TMR, incur significant area and power overheads, which might not always be applicable due to power restrictions. Therefore, we investigate alternative heterogeneous reliability modes t…
▽ More
Reliability has emerged as a key topic of interest for researchers around the world to detect and/or mitigate the side effects of decreasing transistor sizes, such as soft errors. Traditional solutions, like DMR and TMR, incur significant area and power overheads, which might not always be applicable due to power restrictions. Therefore, we investigate alternative heterogeneous reliability modes that can be activated at run-time based on the system requirements, while reducing the power and area overheads of the processor. Our heterogeneous reliability modes are successful in reducing the processor vulnerability by 87% on average, with area and power overheads of 10% and 43%, respectively. To further enhance the design space of heterogeneous reliability, we investigate combinations of efficient compression techniques like Distributed Multi-threaded Checkpointing, Hash-based Incremental Checkpointing, and GNU zip, to reduce the storage requirements of data that are backed-up at an application checkpoint. We have successfully reduced checkpoint sizes by a factor ~6x by combining various state compression techniques. We use gem5 to implement and simulate the state compression techniques and the heterogeneous reliability modes discussed in this paper.
△ Less
Submitted 12 July, 2019; v1 submitted 19 November, 2018;
originally announced November 2018.
-
Malware Detection Approach for Android systems Using System Call Logs
Authors:
Sanya Chaba,
Rahul Kumar,
Rohan Pant,
Mayank Dave
Abstract:
Static detection technologies based on signature-based approaches that are widely used in Android platform to detect malicious applications. It can accurately detect malware by extracting signatures from test data and then comparing the test data with the signature samples of virus and benign samples. However, this method is generally unable to detect unknown malware applications. This is because,…
▽ More
Static detection technologies based on signature-based approaches that are widely used in Android platform to detect malicious applications. It can accurately detect malware by extracting signatures from test data and then comparing the test data with the signature samples of virus and benign samples. However, this method is generally unable to detect unknown malware applications. This is because, sometimes, the machine code can be converted into assembly code, which can be easily read and understood by humans. Furthuremore, the attacker can then make sense of the assembly instructions and understand the functioning of the program from the same. Therefore we focus on observing the behaviour of the malicious software while it is actually running on a host system. The dynamic behaviours of an application are conducted by the system call sequences at the end. Hence, we observe the system call log of each application, use the same for the construction of our dataset, and finally use this dataset to classify an unknown application as malicious or benign.
△ Less
Submitted 25 September, 2017;
originally announced September 2017.
-
A Novel Progressive Multi-label Classifier for Classincremental Data
Authors:
Mihika Dave,
Sahil Tapiawala,
Meng Joo Er,
Rajasekar Venkatesan
Abstract:
In this paper, a progressive learning algorithm for multi-label classification to learn new labels while retaining the knowledge of previous labels is designed. New output neurons corresponding to new labels are added and the neural network connections and parameters are automatically restructured as if the label has been introduced from the beginning. This work is the first of the kind in multi-l…
▽ More
In this paper, a progressive learning algorithm for multi-label classification to learn new labels while retaining the knowledge of previous labels is designed. New output neurons corresponding to new labels are added and the neural network connections and parameters are automatically restructured as if the label has been introduced from the beginning. This work is the first of the kind in multi-label classifier for class-incremental learning. It is useful for real-world applications such as robotics where streaming data are available and the number of labels is often unknown. Based on the Extreme Learning Machine framework, a novel universal classifier with plug and play capabilities for progressive multi-label classification is developed. Experimental results on various benchmark synthetic and real datasets validate the efficiency and effectiveness of our proposed algorithm.
△ Less
Submitted 22 September, 2016;
originally announced September 2016.
-
A novel online multi-label classifier for high-speed streaming data applications
Authors:
Rajasekar Venkatesan,
Meng Joo Er,
Mihika Dave,
Mahardhika Pratama,
Shiqian Wu
Abstract:
In this paper, a high-speed online neural network classifier based on extreme learning machines for multi-label classification is proposed. In multi-label classification, each of the input data sample belongs to one or more than one of the target labels. The traditional binary and multi-class classification where each sample belongs to only one target class forms the subset of multi-label classifi…
▽ More
In this paper, a high-speed online neural network classifier based on extreme learning machines for multi-label classification is proposed. In multi-label classification, each of the input data sample belongs to one or more than one of the target labels. The traditional binary and multi-class classification where each sample belongs to only one target class forms the subset of multi-label classification. Multi-label classification problems are far more complex than binary and multi-class classification problems, as both the number of target labels and each of the target labels corresponding to each of the input samples are to be identified. The proposed work exploits the high-speed nature of the extreme learning machines to achieve real-time multi-label classification of streaming data. A new threshold-based online sequential learning algorithm is proposed for high speed and streaming data classification of multi-label problems. The proposed method is experimented with six different datasets from different application domains such as multimedia, text, and biology. The hamming loss, accuracy, training time and testing time of the proposed technique is compared with nine different state-of-the-art methods. Experimental studies shows that the proposed technique outperforms the existing multi-label classifiers in terms of performance and speed.
△ Less
Submitted 31 August, 2016;
originally announced September 2016.
-
Malicious User Detection in Spectrum Sensing for WRAN Using Different Outliers Detection Techniques
Authors:
Manish B Dave,
Mitesh B Nakrani
Abstract:
In cognitive radio it is of prime importance that the presence of Primary Users (PU) is detected correctly at each of the time. In order to do so the help from all present Secondary Users (SU) is taken and such a taken is known as co-operative spectrum sensing. Ideally it is assumed that all the secondary users give the correct result to the control center. But there are certain conditions under w…
▽ More
In cognitive radio it is of prime importance that the presence of Primary Users (PU) is detected correctly at each of the time. In order to do so the help from all present Secondary Users (SU) is taken and such a taken is known as co-operative spectrum sensing. Ideally it is assumed that all the secondary users give the correct result to the control center. But there are certain conditions under which the secondary users deliberately forward wrong result to the control center so as to degrade the performance of the cognitive network. In this paper we study the different techniques for detecting the malicious users or outliers. We take into consideration practical environmental condition such that the received signal of the secondary users is made to undergo fading and noise is also introduced in the signal. We further go on to examine each of the outlier detector techniques and find out the most suitable at various instants.
△ Less
Submitted 12 May, 2014;
originally announced May 2014.
-
Optimal number of users in Co-operative spectrum sensing in WRAN using Cyclo-Stationary Detector
Authors:
Manish B Dave
Abstract:
Cognitive radio allows unlicensed users to access licensed frequency bands through dynamic spectrum access so as to reduce spectrum scarcity. This requires intelligent spectrum sensing techniques. This paper investigates the use of cyclo-stationary detector and performance evaluation for Digital Video Broadcast-Terrestrial (DVB-T) signals. Generally, DVB-T is specified in IEEE 802.22 standard in V…
▽ More
Cognitive radio allows unlicensed users to access licensed frequency bands through dynamic spectrum access so as to reduce spectrum scarcity. This requires intelligent spectrum sensing techniques. This paper investigates the use of cyclo-stationary detector and performance evaluation for Digital Video Broadcast-Terrestrial (DVB-T) signals. Generally, DVB-T is specified in IEEE 802.22 standard in VHF and UHF TV broadcasting spectrum. Simulations results show that implementing co-operative spectrum sensing help in better utilization of resources. The paper further proposes to find number of optimal users in a scenario to optimize the detection probability and makes use of the particle swarm optimization (PSO) technique to find an optimum value of threshold.
△ Less
Submitted 13 March, 2014;
originally announced March 2014.
-
Mobile Agent as an Approach to Improve QoS in Vehicular Ad Hoc Network
Authors:
Rakesh Kumar,
Mayank Dave
Abstract:
Vehicular traffic is a foremost problem in modern cities. Huge amount of time and resources are wasted while traveling due to traffic congestion. With the introduction of sophisticated traffic management systems, such as those incorporating dynamic traffic assignments, more stringent demands are being placed upon the available real time traffic data. In this paper we have proposed mobile agent as…
▽ More
Vehicular traffic is a foremost problem in modern cities. Huge amount of time and resources are wasted while traveling due to traffic congestion. With the introduction of sophisticated traffic management systems, such as those incorporating dynamic traffic assignments, more stringent demands are being placed upon the available real time traffic data. In this paper we have proposed mobile agent as a mechanism to handle the traffic problem on road. Mobile software agents can be used to provide the better QoS (Quality of Service) in vehicular ad hoc network to improve the safety application and driver comfort.
△ Less
Submitted 10 August, 2011;
originally announced August 2011.
-
A Comparative Study of Various Routing Protocols in VANET
Authors:
Rakesh Kumar,
Mayank Dave
Abstract:
Vehicular Ad Hoc Networks (VANET) is a subclass of Mobile ad hoc networks which provides a distinguished approach for Intelligent Transport System (ITS). The survey of routing protocols in VANET is important and necessary for smart ITS. This paper discusses the advantages / disadvantages and the applications of various routing protocols for vehicular ad hoc networks. It explores the motivation beh…
▽ More
Vehicular Ad Hoc Networks (VANET) is a subclass of Mobile ad hoc networks which provides a distinguished approach for Intelligent Transport System (ITS). The survey of routing protocols in VANET is important and necessary for smart ITS. This paper discusses the advantages / disadvantages and the applications of various routing protocols for vehicular ad hoc networks. It explores the motivation behind the designed, and traces the evolution of these routing protocols. F inally the paper concludes by a tabular comparison of the various routing protocols for VANET.
△ Less
Submitted 10 August, 2011;
originally announced August 2011.
