DLM: Delayed location management in network mobility (NEMO)-based public transportation systems☆☆,

https://doi.org/10.1016/j.jnca.2016.12.010Get rights and content

Abstract

Network mobility basic support (NEMO-BS) supports efficient group mobility. However, when NEMO-BS is applied to public transportation systems where mobile nodes (MNs) frequently get in/off the public transportation, significant signaling overhead owing to frequent and unnecessary binding updates can occur. To address this problem, we propose a delayed location management (DLM) scheme where an MN postpones its binding update for a pre-defined timer to mitigate the binding update overhead. To evaluate the performance of DLM, we develop an analytical model for the binding update cost and the packet delivery cost during the boarding time. Also, a timer selection algorithm is proposed to optimize the performance of DLM. Evaluation results demonstrate that DLM can reduce the binding update cost and packet delivery cost in a balanced manner by choosing an appropriate timer.

Introduction

Network mobility basic support (NEMO-BS) is a mobility support protocol where a collective mobility of multiple mobile nodes (MNs) is handled as a single unit (Devarapalli et al., 2005, Lee et al., 2012). When MNs are connected to a mobile network (MONET), a mobile router (MR) broadcasts a router advertisement (RA) message with its mobile network prefix (MNP) and then MNs configure their care of addresses (CoAs) based on the MR's MNP. After that, MNs conduct binding updates to their home agents (HAs). Then, when the MONET moves to a new access router (AR), only MR conducts the binding update to its HA while MNs in the MONET do not need to execute any binding updates.

However, when NEMO-BS is applied to a public transportation, unnecessary signaling overhead due to binding updates can occur since MNs frequently get in/off the public transportation. Specifically, when an MN gets off before the public transportation moves to another AR (i.e., an MN has a short boarding time), the binding update for MN's CoA based on the MR's MNP can be unnecessary. Fig. 1 shows an example of the unnecessary binding update. When an MN gets in a public transportation (Step 1 in Fig. 1), the MN configures its CoA based on the MR's MNP and conducts a binding update to its HA (Steps 2–3 in Fig. 1). Then, when the public transportation moves to another bus station (Step 4 in Fig. 1), the MN gets off the public transportation (Step 5 in Fig. 1). In this case, the binding update in Step 3 for supporting collective mobility is useless. Note that the distance between two bus stops in local bus service is typically 300400m (Washington Metropolitan Area Transit Authority) and the maximum diameter for one macro-cell is 3 km in urban areas (3GPPBS 3GPP). In such environments, there is non-negligible probability that an MN gets off before the public transportation moves to another AR.

Intuitively, if an MN with short boarding time does not conduct instantly the binding update when the MN gets in the public transportation, such unnecessary binding update can be reduced. Based on this idea, we propose a delayed location management (DLM) scheme where an MN postpones its binding update until a pre-defined timer T expires. In DLM, the mobility of the MN is managed by mobile IPv6 (MIPv6) before the timer expiration. On the other hand, after the timer expiration, the mobility of the MN is handled by the MR. Therefore, the packets to the MN are forwarded through MN's HA, MR's HA, and MR. Also, the MN does not need to conduct any binding update when the public transportation handovers to another AR. Note that we only consider the scenario where the MN resides in the public transportation. That is, the procedure after the MN gets off the public transportation is beyond scope of this paper. To evaluate the performance of DLM, we develop an analytical model for the binding update cost and the packet delivery cost during the MN attachment time.1 By delaying the binding update, inefficient packet routing caused by the NEMO-BS can be diminished2 (i.e., the packet delivery cost can be reduced). However, if the binding update is excessively delayed, all MNs conduct their binding updates individually whenever the public transportation moves across another AR, and therefore the binding update cost increases significantly. To balance the reduced packet delivery cost and the increased binding update cost, we propose a timer selection algorithm based on the developed analytical model. Evaluation results demonstrate that DLM can reduce the binding update cost and packet delivery cost in a balanced manner by choosing an appropriate timer.

The main contribution of this paper is two-fold: 1) DLM can be implemented with different mobility management schemes (i.e., NEMO-BS and MIPv6) by adjusting the timer value. Accordingly, DLM can achieve adaptive performance optimization in network mobility environment; and 2) extensive evaluation results are presented and analyzed under various environments to assess the performance of DLM.

The remainder of this paper is organized as follows. The related works are summarized in Section 2. The detailed operations of NEMO-BS and DLM are described in 3 Background, 4 Delayed location management (DLM), respectively. The performance analysis model and a timer selection algorithm for finding the optimal timer are illustrated in Section 5. Evaluation results and concluding remarks are given in 6 Evaluation results, 7 Conclusion, respectively.

Section snippets

Related works

To improve the performance of NEMO-BS, a number of schemes have been proposed in the literature (Qiang et al., 2014, Kim et al., 2005, Cho et al., 2006, Calderon et al., 2006, Chuang and Lee, 2011, Barman et al., 2015, Kabir et al., 2013, Ernest et al., 2016, Nguyen and Bonnet, 2015, Pack et al., 2009). Qiang et al. (2014) suggested an adaptive route optimization scheme which consists of the mobility transparency sub-scheme and the time saving sub-scheme, and a threshold is introduced to

Background

In this section, binding update and packet delivery procedures of NEMO-BS are introduced.

Delayed location management (DLM)

In this section, we explain the operations of DLM, which are dependent on whether the timer T expires or not. For example, before the timer T expires, the packets destined to the MN are transmitted through only HA_MN and the MN conducts the binding update whenever the public transportation moves to another AR. On the other hand, after the timer expiration, when a CN sends packets to the MN, the packets are transmitted through HA_MN, HA_MR, and MR. Also, since the mobility of the MN is managed

Performance analysis

In this section, we develop an analytical model for the total cost that consists of the binding update cost and the packet delivery cost during the MN attachment time. The MN attachment time represents the period between when an MN gets on and when it gets off the public transportation (i.e., the boarding time of the MN). Also, a timer selection algorithm is introduced to minimize the total cost. Important notations for the analytical model are summarized in Table 1.

Evaluation results

For performance evaluation, we compare DLM against MIPv6 and NEMO-BS. Default parameter settings are summarized in Table 2.

To verify the accuracy of the analytical model, we develop an event-driven simulator using MATLAB and conduct extensive simulations. Table 3 shows the analytical and simulation results where the timer is selected by the timer selection algorithm. From Table 3, it can be shown that the analytical results are consistent with the simulation results since the difference between

Conclusion

To reduce unnecessary binding update for MN's CoA based on the MR's MNP, we have proposed a delayed location management (DLM) scheme where an MN postpones its binding update until a pre-defined timer T expires. To optimize the performance of DLM, we devised a timer selection algorithm and the optimal timer is adaptively set to an appropriate value. Evaluation results demonstrate that DLM outperforms existing schemes under various environments. In our future work, we will evaluate the proposed

References (20)

  • P. Ernest et al.

    Design and performance evaluation of distributed mobility management schemes for network mobility

    J. Netw. Comput. Appl. (JNCA)

    (2016)
  • 3GPPBS 3GPP, 3rd Generation Partnership Project; Technical Specification Group Services and System Aspects;...
  • Barman, S., Ghosh, A., Biswas, S., 2015. A transparent tree root identification scheme to support route-optimization...
  • M. Calderon et al.

    Design and experimental evaluation of a route optimization solution for NEMO

    IEEE J. Sel. Areas Commun. (JSAC)

    (2006)
  • H. Cho et al.

    Route optimization using tree information option for nested mobile networks

    IEEE J. Sel. Areas Commun. (JSAC)

    (2006)
  • M. Chuang et al.

    DRO domain-based route optimization scheme for nested mobile networks

    EURASIP J. Wirel. Commun. Netw.

    (2011)
  • V. Devarapalli et al.

    Network mobility (NEMO) basic support protocol

    RFC

    (2005)
  • B. Haberman et al.

    IPv6 router advertisement flags option

    IETF RFC 5175

    (2008)
  • Md.H. Kabir et al.

    Efficient route optimization scheme for nested-NEMO

    J. Netw. Comput. Appl. (JNCA)

    (2013)
  • Kim, H., Kim, G., Kim, C., 2005. S-RO: Simple route optimization scheme with NEMO transparency, In: Proceedings...
There are more references available in the full text version of this article.

Cited by (4)

  • Popularity-aware mobile content management: When is knowing less better?

    2020, Computer Networks
    Citation Excerpt :

    Ernest et al. [7] investigate a distributed solution for network mobility, focusing on the impact of different levels of network nesting (where nodes in the same level move as a single entity). Motivated by the overhead of binding updates in scenarios with high mobility, Ko et al. [8] propose a scheme that delays binding updates, investigating the influence of the binding update cost and the traffic packet delivery cost due to old bindings. The above works consider a centralized control.

☆☆

This work was supported by the R&D program of MOTIE/KEIT [10051306, Development of Vehicular Cloud-based Dynamic Security Framework for Internet of Vehicles (IoV) Services] and National Research Foundation of Korea Grant funded by the Korean Government (NRF-2014R1A2A1A12066986).

A preliminary version of this paper was presented at the 12th EAI International Conference on Heterogeneous Networking for Quality, Reliability, Security and Robustness (QShine) 2016, July 2016 (Ko et al., 2016).

View full text