TABLE OF CONTENTS In Section 5, we have shown that choosing the optimal values of the scheduling parameters is the key to achieve the optimal capacity-delay tradeoff. In this section, we will show that deviating from these optimal values will lead to suboptimal capacity-delay tradeoffs. In particular, we will identify the limiting factors in the existing schemes in [3] and [4] by comparing the optimal values of scheduling parameters in Section 5.1 with those used by the existing schemes. Our model in Section 4 can be extended to study the upper bounds on the capacity-delay tradeoff when one imposes additional restrictive assumptions that correspond to these limiting factors. We will see that these new upper bounds are inferior to the capacity-delay tradeoff reported in Sections 4 and 5. The existing schemes of [3] and [4] in fact achieve capacity-delay tradeoffs that are close to the respective upper bounds. These results will give us new insights on which schemes to use under different conditions.
The scheme by Neely and Modiano [3] divides the unit square into n cells each of area 1 /n. A mobile relay will forward messages to the destination only when they both reside in the same cell. Hence, the distance from the last mobile relay to the destination, l b, is on average on the order of O(1/ 
), regardless of the delay constraints. However, we have shown...
