Wenliang Du

A pairwise key predistribution scheme for wireless sensor networks

To achieve security in wireless sensor networks, it is important to be able to encrypt and authenticate messages sent between sensor nodes. Before doing so, keys for performing encryption and authentication must be agreed upon by the communicating parties. Due to resource constraints, however, achieving key agreement in wireless sensor networks is nontrivial. Many key agreement schemes used in general networks, such as Diffie-Hellman and other public-key based schemes, are not suitable for wireless sensor networks due to the limited computational abilities of the sensor nodes. Predistribution of secret keys for all pairs of nodes is not viable due to the large amount of memory this requires when the network size is large.In this paper, we provide a framework in which to study the security of key predistribution schemes, propose a new key predistribution scheme which substantially improves the resilience of the network compared to previous schemes, and give an in-depth analysis of our scheme in terms of network resilience and associated overhead. Our scheme exhibits a nice threshold property: when the number of compromised nodes is less than the threshold, the probability that communications between any additional nodes are compromised is close to zero. This desirable property lowers the initial payoff of smaller-scale network breaches to an adversary, and makes it necessary for the adversary to attack a large fraction of the network before it can achieve any significant gain.

A key management scheme for wireless sensor networks using deployment knowledge

To achieve security in wireless sensor networks, it is important to he able to encrypt messages sent among sensor nodes. Keys for encryption purposes must he agreed upon by communicating nodes. Due to resource constraints, achieving such key agreement in wireless sensor networks is nontrivial. Many key agreement schemes used in general networks, such as Diffie-Hellman and public-key based schemes, are not suitable for wireless sensor networks. Pre-distribution of secret keys for all pairs of nodes is not viable due to the large amount of memory used when the network size is large. Recently, a random key pre-distribution scheme and its improvements have been proposed. A common assumption made by these random key pre-distribution schemes is that no deployment knowledge is available. Noticing that in many practical scenarios, certain deployment knowledge may be available a priori, we propose a novel random key pre-distribution scheme that exploits deployment knowledge and avoids unnecessary key assignments. We show that the performance (including connectivity, memory usage, and network resilience against node capture) of sensor networks can he substantially improved with the use of our proposed scheme. The scheme and its detailed performance evaluation are presented in this paper.

Using randomized response techniques for privacy-preserving data mining

Privacy is an important issue in data mining and knowledge discovery. In this paper, we propose to use the randomized response techniques to conduct the data mining computation. Specially, we present a method to build decision tree classifiers from the disguised data. We conduct experiments to compare the accuracy of our decision tree with the one built from the original undisguised data. Our results show that although the data are disguised, our method can still achieve fairly high accuracy. We also show how the parameter used in the randomized response techniques affects the accuracy of the results.

DroidAPIMiner: Mining API-Level Features for Robust Malware Detection in Android

The increasing popularity of Android apps makes them the target of malware authors. To defend against this severe increase of Android malwares and help users make a better evaluation of apps at install time, several approaches have been proposed. However, most of these solutions suffer from some shortcomings; computationally expensive, not general or not robust enough. In this paper, we aim to mitigate Android malware installation through providing robust and lightweight classifiers. We have conducted a thorough analysis to extract relevant features to malware behavior captured at API level, and evaluated different classifiers using the generated feature set. Our results show that we are able to achieve an accuracy as high as 99% and a false positive rate as low as 2.2% using KNN classifier.

Mitigating DoS attacks against broadcast authentication in wireless sensor networks

Broadcast authentication is a critical security service in wireless sensor networks. There are two general approaches for broadcast authentication in wireless sensor networks: digital signatures and μTESLA-based techniques. However, both signature-based and μTESLA-based broadcast authentication are vulnerable to Denial of Services (DoS) attacks: An attacker can inject bogus broadcast packets to force sensor nodes to perform expensive signature verifications (in case of signature-based broadcast authentication) or packet forwarding (in case of μTESLA-based broadcast authentication), thus exhausting their limited battery power. This paper presents an efficient mechanism called message-specific puzzle to mitigate such DoS attacks. In addition to signature-based or μTESLA-based broadcast authentication, this approach adds a weak authenticator in each broadcast packet, which can be efficiently verified by a regular sensor node, but takes a computationally powerful attacker a substantial amount of time to forge. Upon receiving a broadcast packet, each sensor node first verifies the weak authenticator, and performs the expensive signature verification (in signature-based broadcast authentication) or packet forwarding (in μTESLA-based broadcast authentication) only when the weak authenticator is valid. A weak authenticator cannot be precomputed without a non-reusable (or short-lived) key disclosed only in a valid packet. Even if an attacker has intensive computational resources to forge one or more weak authenticators, it is difficult to reuse these forged weak authenticators. Thus, this weak authentication mechanism substantially increases the difficulty of launching successful DoS attacks against signature-based or μTESLA-based broadcast authentication. A limitation of this approach is that it requires a powerful sender and introduces sender-side delay. This article also reports an implementation of the proposed techniques on TinyOS, as well as initial experimental evaluation in a network of MICAz motes.

Privacy-preserving cooperative statistical analysis

The growth of the Internet opens up tremendous opportunities for cooperative computation, where the answer depends on the private inputs of separate entities. Sometimes these computations may occur between mutually untrusting entities. The problem is trivial if the context allows the conduct of these computations by a trusted entity that would know the inputs from all the participants; however if the context disallows this then the techniques of secure multiparty computation become very relevant and can provide useful solutions. Statistical analysis is a widely used computation in real life, but the known methods usually require one to know the whole data set; little work has been conducted to investigate how statistical analysis could be performed in a cooperative environment, where the participants want to conduct statistical analysis on the joint data set, but each participant is concerned about the confidentiality of its own data. We have developed protocols for conducting the statistical analysis in such a cooperative environment based on a data perturbation technique and cryptography primitives.

An efficient scheme for authenticating public keys in sensor networks

With the advance of technology, Public Key Cryptography (PKC) will sooner or later be widely used in wireless sensor networks. Recently, it has been shown that the performance of some public-key algorithms, such as Elliptic Curve Cryptography (ECC), is already close to being practical on sensor nodes. However, the energy consumption of PKC is still expensive, especially compared to symmetric-key algorithms. To maximize the lifetime of batteries, we should minimize the use of PKC whenever possible in sensor networks.This paper investigates how to replace one of the important PKC operations--the public key authentication--with symmetric key operations that are much more efficient. Public key authentication is to verify the authenticity of another partys public key to make sure that the public key is really owned by the person it is claimed to belong to. In PKC, this operation involves an expensive signature verification on a certificate. We propose an efficient alternative that uses one-way hash function only. Our scheme uses all sensors public keys to construct a forest of Merkle trees of different heights. By optimally selecting the height of each tree, we can minimize the computation and communication costs. The performance of our scheme is evaluated in the paper.

A key predistribution scheme for sensor networks using deployment knowledge

To achieve security in wireless sensor networks, it is important to be able to encrypt messages sent among sensor nodes. Keys for encryption purposes must be agreed upon by communicating nodes. Due to resource constraints, achieving such key agreement in wireless sensor networks is nontrivial. Many key agreement schemes used in general networks, such as Diffie-Hellman and public-key-based schemes, are not suitable for wireless sensor networks. Predistribution of secret keys for all pairs of nodes is not viable due to the large amount of memory used when the network size is large. Recently, a random key predistribution scheme and its improvements have been proposed. A common assumption made by these random key predistribution schemes is that no deployment knowledge is available. Noticing that, in many practical scenarios, certain deployment knowledge may be available a priori, we propose a novel random key predistribution scheme that exploits deployment knowledge and avoids unnecessary key assignments. We show that the performance (including connectivity, memory usage, and network resilience against node capture) of sensor networks can be substantially improved with the use of our proposed scheme. The scheme and its detailed performance evaluation are presented in this paper.

A witness-based approach for data fusion assurance in wireless sensor networks

In wireless sensor networks, sensor nodes are spread randomly over the coverage area to collect information of interest. Data fusion is used to process these collected information before they are sent to the base station, the observer of the sensor network. We study the security of the data fusion process in this work. In particular, we propose a witness-based solution to assure the validation of the data sent from data fusion nodes to the base station. We also present the theoretical analysis for the overhead associated with the mechanism, which indicates that even in an extremely harsh environment the overhead is low for the proposed mechanism.

Secure Multi-party Computational Geometry

The general secure multi-party computation problem is when multiple parties (say, Alice and Bob) each have private data (respectively, a and b) and seek to compute some function f(a, b) without revealing to each other anything unintended (i.e., anything other than what can be inferred from knowing f(a, b)). It is well known that, in theory, the general secure multi-party computation problem is solvable using circuit evaluation protocols. While this approach is appealing in its generality, the communication complexity of the resulting protocols depend on the size of the circuit that expresses the functionality to be computed. As Goldreich has recently pointed out [6], using the solutions derived from these general results to solve specific problems can be impractical; problem-specific solutions should be developed, for efficiency reasons. This paper is a first step in this direction for the area of computational geometry. We give simple solutions to some specific geometric problems, and in doing so we develop some building blocks that we believe will be useful in the solution of other geometric and combinatorial problems as well.