A named data networking prediction-based mobility solution in space–air–terrestrial networks

Abstract

With the rapid development of satellite constellations and mobile devices, multimedia-content transfer in space–air–terrestrial networks (SATNs) has become increasingly popular. Governed by the current internet protocol-based architecture and the distance between the space and the ground, real-time transmissions of multimedia encounter an enormous challenge. Recently, named data networking has been considered to be a promising approach for solving this problem in SATN scenarios. However, a new issue, producer and consumer mobility management, arises in this approach. In this paper, we propose a location prediction-based scheme to reduce the round-trip time during the handover process and further enhance the system throughput, as well as to enable the producer and consumer to simultaneously move from an old access point (AP) to a new AP. Moreover, the proposed scheme can be applied in two kinds of SATN scenarios. Finally, the effectiveness of the proposed scheme is verified by simulations.

Introduction

Recently, with the rapid development of satellite constellations such as O3b, OneWeb, and Starlink Internet built by SpaceX, space–air–terrestrial networks (SATNs) have attracted increasing attention. These networks operate on medium and low orbits that can greatly reduce the round-trip transmission delay compared with the geosynchronous equatorial orbit satellite. Moreover, megasatellite constellations with hundreds/thousands of small satellites (such as low earth orbit (LEO) satellites) can achieve a wide range of coverage, greatly reducing hardware costs. In the upcoming Space 2.0 era [1], several space nodes are equipped with powerful communication links to provide worldwide service and play the roles of content producers and/or content consumers.

Moreover, with the rapid development of mobile devices, such as smart mobile phones, pads, and unmanned aerial vehicles (UAVs), multimedia content has become the mainstream content transmitted over networks and occupies an increasing proportion of network bandwidth. The current internet transmission control protocol (TCP)/Internet protocol (IP) architecture is based on host–client communications, where the address must be known when accessing the contents. In a case where a UAV group transfers the media data simultaneously through the same access point (AP), which is a common scenario in the surveying and mapping fields, a traffic jam is likely to occur. As the UAV energy issue is not imperative, Refs. [2] and [3] have developed some solutions. On the other hand, it requires putting more effort into the traffic jam issue to improve the performance. Recently, some solutions have been proposed for a future Internet architecture to fix this issue [4], [5], [6]. Information-centric networking (ICN) has been the most common candidate approach proposed in recent work. Compared with the classic TCP/IP architecture, ICN is content-based since users focus on the data content and their names rather than the address of the contents. Moreover, the routers in TCP/IP systems transfer the packages by the host, and the ICN routers forward the packages by the content name. Named data networking (NDN) is an ICN architecture that is currently very active and agile.

Generally, there are two roles in the NDN infrastructure: content producers and content consumers. Moreover, two types of transferred packages, interest and data, are applied to the communications between both roles. When a consumer requests the content, an interesting package will be broadcast to the Internet through the connected routers. Then, the routers in NDN transfer interest to the content producer. Once the data package for the interest is generated, it will be transferred to the consumer through the routers from the backtrace. Commonly, the interest and the content data transferred by the routers will be stored for future uses. Once the router receives the same interest package, it can reply by using the cached content immediately, leading to greatly improved efficiency.

As shown in Fig. 1, there are three tables at an NDN router for interest and data package forwarding, that is, the pending interest table (PIT), the forwarding information base (FIB), and the content store (CS). The PIT is used for the transmissions of the arrived data packets by recording each forwarded interest as well as the corresponding incoming face (port) of the interest [7]. The FIB stores multiple output faces mapped to a name prefix to forward each interest packet to the producer [7], while the CS stores the temporarily forwarded data packages by the router.

Research in NDN has mainly focused on the naming resolution, forwarding strategy, routing, security, and mobility. Mobility management is becoming more salient because of the transformation of devices into smart gadgets that rapidly create and share content in real-time [8]. Therefore, this research is has attracted intense interest. Based on the NDN design, consumer mobility is intrinsically supported. When the consumer changes the position from location A to location B, sends the interest repeatedly, and waits for the data back in B, the router in A will drop the PIT item until the interest lifetime is over. There is no need to notify the network of the change in the consumer’s location since the link will recover when the consumer resends the interest at a new position, which may require more trip time in data receiving. By contrast, producer mobility is not defined in NDNs and is still a changing task. The reason is that producer mobility may cause the update of the package forwarding plane, particularly in some high-speed traveling scenarios such as SATN.

To date, several works have been carried out to manage both consumer mobility and producer mobility, but only a small number of these results consider mobility in SATNs. The authors of [9] built a space–terrestrial network by using an anchorless scheme to address the producer mobility issue. However, the focus of this work is the router algorithm that finds the shortest path of the multilayered satellite networks, not the consumer mobility. The authors of [8] proposed an anchorless producer mobility management solution that focuses on real-time communications in wireless networks, and the scene does not cover the SATN and consumer mobility.

In this paper, we build a future SATN model illustrated in Fig. 2. In this scenario, UAVs and civil aircraft (CAs) act as both consumers and producers, resulting in continuous position movement at high speed. Simultaneously, the satellites, which play the role of a relay point of the network to be connected by UAVs or CAs, operate periodically along their orbits at a high speed. There are security considerations due to the open environment, e.g., Ref. [10] provided a stochastic sampling-based cache learning strategy to solve the security of multi-relay networks. On the other hand, UAVs and CAs require real-time transmissions because they need to upload the collected video and the status information of CA to facilitate the passengers in aircraft watching the video provided by producers in terrestrial networks. To solve the above-described issues, we propose a prediction-based scheme. The key points of our work are as follows.

(1) For UAV and CA as consumers, our method predicts the new AP and establishes the link between old AP and new AP that is predicted by the algorithm, and then, the PIT is forwarded to the new AP to reduce the RTT during the handover. If data are received, the data is sent to the predicted AP.

(2) For UAV and CA as producers, the proposed algorithm connects the old AP and new AP once the prediction event occurred. If the interest packages are received by the old AP, the interest redirect (interest RDRCT) is sent to the new AP. This can reduce the RTT while UAV and CA change their locations.

(3) We conduct evaluation experiments to demonstrate the advantages of our algorithm. The results show that the proposed scheme achieves shorter RTT and enhances the throughput during handover.

The remainder of this paper is organized as follows. Section 2 introduces the related work, particularly the consumer mobility methods and producer mobility solutions. Section 3 first introduces the normal procedure for considering consumer and producer mobility. Second, we describe the basic logic of location detection and AP prediction for the UAV and CA scenarios. Third, we propose our consumer and producer mobility solution based on AP prediction for both UAV and CA scenes. Then, Section 4 describes the experimental environment and the evaluation results in both scenarios. Finally, Section 5 summarizes the paper and describes future work.