### Table 4. Key management schemes for mobile ad hoc networks.

in Abstract

"... In PAGE 31: ... By fabricating and forwarding route error messages an attacker can try to disrupt the operation of existing routes, not only breaking connectivity but also creating additional routing overhead in the network as a result of legitimate nodes trying to establish alternative paths. In Table4 we present the route maintenance characteristics of the protocols we have analyzed. The solution adopted by most of the analyzed protocols requires the signing of the complete error message by the node that generates or forwards it.... In PAGE 33: ... In this section we will present several key management solutions that have been specifically proposed to address the challenges of mobile ad hoc networks and discuss their behavior in respect to mobility patterns and operational requirements. Table4 summarizes the results of our analysis. Table 4.... ..."

### Table 9: Multi-hop Ad-Hoc Network Model with Energy-Efficient Routing Link Met- rics

### Table 4: Comparing the median latency and hop- counts of XOR based overlay paths and the under- lying IP paths for two extremely di erent latency distributions (marked VA and JP in Figure 4). The latencies are in milliseconds.

2003

"... In PAGE 12: ... While this question has been addressed in papers proposing indi- vidual proximity designs, we revisit it here because our tests use a more realistic latency distribution and therefore may be more indicative of real-world performance. As the abso- lute performance depends on the exact latency distribution used, in Table4 we show the median latencies of the various designs for two very di erent latency distributions. As can be seen, the very best options (PNS and PNS+PRS) fare quite well for either distribution.... ..."

Cited by 164

### Table 4: Comparing the median latency and hop- counts of XOR based overlay paths and the under- lying IP paths for two extremely di erent latency distributions (marked VA and JP in Figure 4). The latencies are in milliseconds.

"... In PAGE 11: ... While this question has been addressed in papers proposing indi- vidual proximity designs, we revisit it here because our tests use a more realistic latency distribution and therefore may be more indicative of real-world performance. As the abso- lute performance depends on the exact latency distribution used, in Table4 we show the median latencies of the various designs for two very di erent latency distributions. As can be seen, the very best options (PNS and PNS+PRS) fare quite well for either distribution.... ..."

### Table 4: Comparing the median latency and hop- counts of XOR based overlay paths and the under- lying IP paths for two extremely different latency distributions (marked VA and JP in Figure 4). The latencies are in milliseconds.

"... In PAGE 11: ... While this question has been addressed in papers proposing indi- vidual proximity designs, we revisit it here because our tests use a more realistic latency distribution and therefore may be more indicative of real-world performance. As the abso- lute performance depends on the exact latency distribution used, in Table4 we show the median latencies of the various designs for two very different latency distributions. As can be seen, the very best options (PNS and PNS+PRS) fare quite well for either distribution.... ..."

### Table 4: Comparing the median latency and hop- counts of XOR based overlay paths and the under- lying IP paths for two extremely di erent latency distributions (marked VA and JP in Figure 4). The latencies are in milliseconds.

"... In PAGE 11: ... While this question has been addressed in papers proposing indi- vidual proximity designs, we revisit it here because our tests use a more realistic latency distribution and therefore may be more indicative of real-world performance. As the abso- lute performance depends on the exact latency distribution used, in Table4 we show the median latencies of the various designs for two very di erent latency distributions. As can be seen, the very best options (PNS and PNS+PRS) fare quite well for either distribution.... ..."

### Table 4: Comparing the median latency and hop- counts of XOR based overlay paths and the under- lying IP paths for two extremely different latency distributions (marked VA and JP in Figure 4). The latencies are in milliseconds.

"... In PAGE 11: ... While this question has been addressed in papers proposing indi- vidual proximity designs, we revisit it here because our tests use a more realistic latency distribution and therefore may be more indicative of real-world performance. As the abso- lute performance depends on the exact latency distribution used, in Table4 we show the median latencies of the various designs for two very different latency distributions. As can be seen, the very best options (PNS and PNS+PRS) fare quite well for either distribution.... ..."

"... In PAGE 11: ... While this question has been addressed in papers proposing indi- vidual proximity designs, we revisit it here because our tests use a more realistic latency distribution and therefore may be more indicative of real-world performance. As the abso- lute performance depends on the exact latency distribution used, in Table4 we show the median latencies of the various designs for two very di erent latency distributions. As can be seen, the very best options (PNS and PNS+PRS) fare quite well for either distribution.... ..."

### Table 1: Energy cost in joules for single source broadcasting in a 100-node network.

2001

"... In PAGE 4: ...nodes with two different field sizes. Table1 presents the lower bounds for average energy cost and the average cost with our scheme for broadcasting in 8 different node distributions. The results show that our simple scheme performs well for broadcasting for dense ad hoc networks.... ..."

Cited by 9