LAPEL: Hop Limit Based Adaptive PIT Entry Lifetime for Vehicular Named Data Networks

Abstract

Named data networking (NDN) uses the content name instead of its location within the interest message to locate the content within the network. An interest is broadcasted by the content seeking node over the wireless channel with specific content name and waits for the data message containing the actual content. The broadcast region of an interest is restricted by using the time-to-live (TTL) field of the message. Record of every pending interest is stored in the pending interest table (PIT) for a specific duration, called PIT entry lifetime (PEL). The record is purged when a node either receives the data message or the PEL expires. The size of the PIT has a significant impact on the name prefix search during the interest or data message forwarding process. Additionally, the interest and/or data message loss in the vehicular network is very frequent, which results in a significant PIT size depending on the PEL. Vanilla NDN uses the fixed PEL of $4\text{ s}$ , and the very recent work considers the constant parameters to control the PEL, which fails to reduce the PIT size. Hence, in this paper, we propose a hop limit based adaptive PEL (LAPEL) scheme, which adaptively estimates the PEL at interest forwarding vehicle. LAPEL estimates PEL based on the lower layer parameters and has no constant factor that makes it more adaptive than the previously proposed schemes. Simulation results show that LAPEL has $52\%$ less PEL, reduces $8.4\%$ the PIT size, and achieves $4.5\%$ more interest satisfaction ratio, for any simulated TTL and interest rates.