Power proximity based key management for secure multicast in ad hoc networks

Abstract

As group-oriented services become the focal point of ad hoc network applications, securing the group communications becomes a default requirement. In this paper, we address the problem of group access in secure multicast communications for wireless ad hoc networks. We argue that energy expenditure is a scarce resource for the energy-limited ad hoc network devices and introduce a crosslayer approach for designing energy-efficient, balanced key distribution trees to perform key management. To conserve energy, we incorporate the network topology (node location), the "power proximity" between network nodes and the path loss characteristics of the medium in the key distribution tree design. We develop new algorithms for homogeneous as well as heterogeneous environments and derive their computational complexity. We present simulation studies showing the improvements achieved for three different but common environments of interest, thus illustrating the need for cross-layer design approaches for security in wireless networks. Index Termsad hoc networks, key management, multicast, security, energy efficiency

Energy and Bandwidth-Efficient Key Distribution in Wireless Ad Hoc Networks: A Cross-Layer Approach

“…In both proposals it was shown that using a Logical Key Hierarchy (LKH) such as a d-ary key tree T, reduces member key storage and GC transmissions to O(log d N ). While key trees are minimal structures in terms of member key storage and GC transmissions, not all key trees are energy-efficient [11], [12]. However, we will show that key tree structures designed by incorporating the metrics of m Ave and E Ave , lead to energy and bandwidth-efficient solutions to the KDP for the wireless ad-hoc networks.…”

“…While the key-tree assignment structure provides scalability in terms of member key storage and GC transmissions, it has been shown to be energy inefficient when the network topology is not taken into account [11], [12]. In fact, it was shown in [14] that updating the SEK via unicasting to each member (N -ary key tree) can require less energy to rekey a member deletion, compared to a binary tree, despite the larger number of GC transmissions in the case of the N -ary tree.…”

“…Furthermore, it was shown in [12], that an energy-optimal solution for the average update energy does not imply an optimal solution to the number of rekeying messages. Since energy is, in most cases, a more scarce resource than bandwidth, we focus on finding an energy-efficient solution and show that our scheme is also bandwidth-efficient.…”

“…We then employed VP3 to partition each network into groups of size d ∈ {3, 4, 5, 6} (steps 1−4 of VP3) and evaluated ∆(S i ) for each of the resulting subgroups, using (12) 3 . The histograms in Figure 9 present the percentage of subgroups that showed a ∆(S i ) > 0, for subgroups of different size.…”

On the security of group communication schemes based on symmetric key cryptosystems

“…Although the aforementioned schemes can be directly deployed in wireless settings, a simple-minded adoption may not lead to the desired performance (see, e.g., [28,15]). Fortunately, there have been some interesting investigations that show that these schemes can be adapted (e.g., by taking into account some physical characteristics of ad hoc networks [16,14,13,15]) so that better performance can be achieved.…”