A location management plays an important role in guaranteeing the effective operation of Personal Communication Services (PCS). With the increasing use of mobile devices, the need of Location Management in PCS Networks has emerged as a challenge to call delivery in an efficient manner. In order to facilitate all suitable parameters for the efficient functions of Location registration and call delivery, various approaches are being explored and modified in these days by research community to optimize the existing location management techniques. This paper makes an attempt to shed light on comparative analysis of such location management strategies.
Keywords |
PCS, GSN Network, Location Management, Call Delivery |
INTRODUCTION |
Personal Communications Services (PCS) support mobile devices which are free to travel within the service coverage
area. The network is composed of several functional entities: The Mobile Station (MS) which is carried by the
subscriber; the BSS (Base Station Subsystem) that controls the radio link with the mobile stations; and the MSC (Mobile
Switching Centre) that performs the switching of calls between the mobile users and between mobile and fixed network
users. MSC is also responsible for registration, authentication, location updating, handovers, call routing to a roaming
subscriber etc. In order to effectively locate an MS when a call is initiated, Location Management Schemes are used
to keep track of the locations of the MTs. The current approach for location management requires each MT to report its
location to the network periodically [13]. The location of the called MT is then stored in the databases. When a call is
initiated, a database lookup and paging procedure determines the current location of the called MT. The cellular system
has registered a rapid growth in world communication scenario now-a-days. The standard like GSM has captured
around 78% of the world communication market through its widespread use in European countries, India etc. In GSM,
a mobile unit (MU) is free to move around within the entire area of coverage. The movement is random and therefore
the geographical location of mobile unit is unpredictable. In order to deliver a call to MU, it is necessary to have
information about its current location. Location management are based on a two-level data hierarchy; Home location
register (HLR) and Visitor Location Register (VLR) that involved the call tracking of MTs. HLR contains all the
administrative information of each subscriber registered in the corresponding GSM networks along with the current
location of the mobile. There is logically one HLR per GSM network, although it may be implemented as a distributed
database. VLR contains selected administrative information from the HLR necessary for all control and provision of
the subscribed services, for each mobile currently located in the geographical area is controlled by the VLR. Personal
Communication Service (PCS) networks provide wireless communication services that enable Mobile Terminals (MTs)
to transfer any form of information between any location at any time [1, 2, 3]. |
RELATED WORK |
The network is composed of several functional entities: The Mobile Station (MS) which is carried by the subscriber; the
BSS (Base Station Subsystem) that controls the radio link with the mobile stations; and the MSC (Mobile Switching
Centre) that performs the switching of calls between the mobile users and between mobile and fixed network users.
MSC is also responsible for registration, authentication, location updating, handovers, call routing to a roaming
subscriber etc. In order to effectively locate an MS when a call is initiated, Location Management Schemes are used
to keep track of the locations of the MTs. The current approach for location management requires each MT to report its
location to the network periodically [13]. The provision of unrestricted mobility of mobile units creates a complex
dynamic environment, and the location management component must be able to identify the correct location of a unit
without any noticeable delay. The location management performs three fundamental tasks: (a) location update, (b) The network is composed of several functional entities: The Mobile Station (MS) which is carried by the subscriber; the
BSS (Base Station Subsystem) that controls the radio link with the mobile stations; and the MSC (Mobile Switching
Centre) that performs the switching of calls between the mobile users and between mobile and fixed network users.
MSC is also responsible for registration, authentication, location updating, handovers, call routing to a roaming
subscriber etc. In order to effectively locate an MS when a call is initiated, Location Management Schemes are used
to keep track of the locations of the MTs. The current approach for location management requires each MT to report its
location to the network periodically [13]. The provision of unrestricted mobility of mobile units creates a complex
dynamic environment, and the location management component must be able to identify the correct location of a unit
without any noticeable delay. The location management performs three fundamental tasks: (a) location update, (b) |
To improve the performance of network in respect with the location management and call delivery, various schemes
have been proposed. The problems faced in centralized approach is not only every location request but service
registration is also operated by a single node called HLR that becomes overloaded and gives rise to more problems like
congestion, delay in call delivery etc. In order to overcome such type of problems, various approaches have been
proposed. The Local Anchor (LA) Scheme [4] has been proposed to reduce the signaling traffic by using a local anchor
(a VLR, a mobile user is currently visiting when a user receives a call). |
In Distributed approach, the information about user location is distributed among various nodes in the network.
Therefore, it provides better stability as compared to centralized approach. However, the operations like location
lookup and location updating are performed in a complex way and hence needs high computation cost. |
In Distributed approach, the information about user location is distributed among various nodes in the network.
Therefore, it provides better stability as compared to centralized approach. However, the operations like location
lookup and location updating are performed in a complex way and hence needs high computation cost. |
PROPOSED SOLUTION |
A. Centralized Approach |
The HLR contains the permanent data of the MSs whose primary subscription is within the area. To assist routing
incoming calls MS contains a pointer to VLR. A VLR is associated with a Mobile Switching Center (MSC) in the
networks and contains temporary record for all MSs currently active within the service area of the MSC. The VLR
retrieves information for handling calls to or from a visiting MS. To facilitate the tracking of a moving MS, a PCS
network is partitioned into many Registration Areas (RAs). In a PCS network, there are several HLRs as shown in fig.
1. |
The service area served by an HLR is referred to as Service Area (SA). The SA that is associated with the master
HLR is called the master SA for the MS. When an MS moves to another new SA, the new SA that the MS resides is
called the current SA. |
The associated HLR is called the current HLR for the MS. The two basic operations in location management are
location registration and call delivery. Location registration is the process through which system tracks the locations of
MSs that move in the networks. When an incoming call arrives, the system searches the called MS (the“callee”) by sending a location request message to the HLR of the callee. The HLR determines the VLR of callee and sends a
location message to the associated MSC. Then the MSC sends the polling signals to all the cells in the RA to determine
the cell location of the callee. This searching process is referred to as call delivery. In the existing location management
schemes, only the master HLR is used for an MS even though it may move to another SA associated with another HLR. |
When an MS moves far away from its master HLR, the communication costs for accessing the master HLR for
both location registration and call delivery will be increased dramatically. This problem leads us to consider why
current HLR of MS cannot be used for the location management to improve the system performance. To improve the
performance of network in respect with the location management and call delivery, various schemes have been
proposed. The problems faced in centralized approach is not only every location request but service registration is also
operated by a single node called HLR that becomes overloaded and gives rise to more problems like congestion, delay
in call delivery etc. In order to overcome such type of problems, various approaches have been proposed. The Local
Anchor (LA) Scheme [4] has been proposed to reduce the signaling traffic by using a local anchor (a VLR, a mobile
user is currently visiting when a user receives a call). In this scheme, a VLR close to the user is selected as the local
anchor (LA) for the user. Whenever a user moves from one RA to another, it will perform location update to the LA. A
LA for a mobile will change only when a call request to the mobile arrives; at the same time, the HLR is also updated
via the registration process. When a call request terminating at this user is received by the HLR, the user can be traced
to the LA. The LA scheme avoids update to HLR completely at the expense of the increase in local signaling traffic.
The drawback of this scheme is that when the user keeps moving constantly without receiving any call, the updates to
LA may become costly, a similar bottleneck as the HLR. Another scheme, Per-user Pointer Forwarding Scheme [5]
has been proposed. In this scheme, some updates to the HLR can be avoided by setting up a forwarding pointer from
the previous VLR to the new VLR. When a call request to a mobile user arrives, the PCS network first queries the
user's HLR to determine the VLR, which the user was visiting at the previous location update, then follows a chain of
forwarding pointers to the user's current VLR to find the mobile user. The traffic to the HLR is decreased by using the
pointer chain; however, the penalty is the time delay for tracking the user when a call to the user arrives. The longer the
pointer chain, the less the signaling traffic, the longer the setup delay for finding the user. To avoid long setup delay, a
threshold of the length of the pointer chain is used. The user needs to perform registration to the HLR after the chain
threshold is reached. In order to overcome the drawbacks these two schemes, Two-level pointer forwarding strategy
was proposed [6]. Two kinds of pointers are used in this scheme. Some VLRs are selected as the Mobility Agents
(MA), which will be responsible for location management in a larger area comparing to the RAs and can be
geographically distributed. But the result concluded is that this strategy can significantly reduce the network signaling
traffic for users with low Call to Mobility Ratio (CMR)[5]-[6] without increasing much of the call setup delay. Another
scheme[7] proposed has suggested caching technique for current standards. In existing HLR/VLR scheme, call is
routed through MSC to LR in which callee is located. When a particular MSC receives a large number of calls to a
particular mobile that belongs to a different home system, the signaling and database cost involved in setting up the call
can be significantly reduced by caching the location information at the calling MSC. Each time when a call is
attempted, the cached information is checked first. Since the access time in looking up an entry in the cached memory
is very short (in microseconds), checking the cached information for every call doesn’t affect the performance of the
MSC. In case of cache hit, call is directly routed to serving LR of callee. But in case of cache miss, HLR is needed to
be contacted and call-establishing time will be longer than normal HLR/VLR scheme. Location information of MH is
cached at an MSC if local call to mobility ratio (LCMR) maintained for the MH at the MSC is larger than a threshold
derived from the link and database access cost of the network. LCMR is the ratio between the numbers of calls
originating from an MSC to the number of times the MH changes its service area as seen by that MSC. However, this
caching scheme may have cached invalid location information as MS moves into a new registration area. So a new
location management strategy based on the centralized database architecture with HLR/VLR concept has been
proposed. The basic idea is to reduce the cost of call delivery using location information cached at MSC/VLR. This
scheme can’t cache invalid location information. There is another scheme, new caching technique [8] proposed as
announcing the location of MS to MSC/VLR or invalidating the location information by MSC/VLR. That is, cached
location information, a miss can’t occur. Therefore, the call setup delay for this proposed scheme is always shorter
than, or equal to that of the current method. This scheme minimizes the total signaling cost on location management by
reducing cost of call delivery more than increased the cost of the location update. An MS has two memory sets, a
candidate and an update set. The candidate set includes the ID of the MSC/VLR that initiates a call for the MS and time
at which call is delivered to compute the call frequency. The candidate set has a limited memory size. If the candidate
set is full, an old entry is replaced by a new entry on LRU (Least Recently used) policy. However, if CMR (Call mobility Ratio) is low i.e. if the mobility is high compared to call request, the scheme may not be sufficient for the
better performance. |
Therefore, another scheme [9], fVLR has come forward to somehow resolve the issue. A mobile user moves anywhere
in mobile networks and the location registration for call tracking is always needed. These registration operations cause
high network traffics. In this scheme, a new simple mobility management is done by registering frequently dialed
mobile phone numbers in a database, called fVLR. Using the fVLR scheme, cost of maintaining location of mobile
users is being reduced. As per this scheme, when set up the call path between mobile users, first of all the VLR of the
caller queries the callee’s fVLR database for searching the location of the callee instead of requesting to HLR of the
callee. The proposed fVLR database is stored or updated whenever mobile phone numbers are dialed by mobile users,
that is, fVLR can be inserted by frequent dialing or deleted by no dialing during a certain period. Nevertheless, this
scheme is feasible for only those users whose mobility is frequent. |
B. Centralized Approach for Personal Communication Services Networks |
As the number of PCS subscribers increase, the system overheads involved with the location management will increase
beyond the capacity of the current network design. Methods for reducing the overheads are critically important for the
design and implementation of PCS networks. A novel dynamic HLR location management scheme, DHLR (dynamic
home location register)[10]for PCS networks has been introduced in which a dynamic copy of location information of
an MS is made in the current (nearest) HLR which can be accessed for location management. In the proposed dynamic
HLR management scheme, when an MS moves from its master SA (Service area) to a new SA, the new current HLR
obtains a copy of the data about the MS from the old current HLR and sends an acknowledgment message to the old
current HLR. An MS can always use the location data in its nearest (current) HLR for performing location registration
and call delivery. This scheme concludes that the proposed method can reduce the total costs of location registration
and call delivery comparing to the existing location management scheme significantly from about 20% to even 70%.
Another scheme to reduce the cost of maintaining the location of users has also been proposed [11]. In this scheme, a
VLR among group of several VLRs is chosen simply for location management. The chosen VLR so called rVLR-B
(that is representative VLR for broadcasting) is registered as the representative VLR and performs broadcasting for
searching mobile users. This scheme reduces the network traffic and involves less registration operations to some
extent. This scheme proposes to manage the representative VLR of several VLRs and registers mobile users’ location.
When set up the call path between mobile users, the VLR of the caller queries callee’s rVLR for searching the location
of callee instead of requesting to VLR of callee. And then, rVLR broadcasts the callee’s location to all VLR of the
region concurrently. Location registration is only performed when a mobile user visits a new rVLR network area from
current area. The selection of rVLR is quite challenging in this scheme for a huge network. |
The centralized approaches discussed above are purely devoted to yield better methods to reduce location management
cost and to improve call delivery in all possible circumstances. |
CONCLUSION AND FUTURE WORK |
As per the above discussion about the centralized location management approaches, it is clear that location
management and call delivery is performed on the basis of HLR and VLR |
The approaches reviewed in the above section try to optimize the location management by making
variations/modifications in the existing HLR/VLR architecture. However, a well-optimized, practically deployable
solution for change in existing GSM architecture is still awaited. This keeps the area of mobile location management as
a fertile research field and we hope to witness more optimized solutions related to this field in near future. |
Figures at a glance |
|
Figure 1 |
|
References
- Jie Li, Yi Pan, Yang Xiao, âÃâ¬ÃËA Dynamic HLR Location Management Scheme for PCS NetworksâÃâ¬Ãâ¢, IEEE INFOCOM 2004
- Y. Fang, Chlamtac and Y. Lin, âÃâ¬ÃËPortable movement modeling for PCS NetworksâÃâ¬Ãâ¢, IEEE Transactions on Vehicular Technology, vol.49, no.4, pp.1356-1363, July 2000.3.
- Y. Fang, Chlamtac and H. Fei, âÃâ¬ÃËAnalytical results for optimal choice of location update interval for mobility database failure restoration in PCS networksâÃâ¬Ãâ¢, IEEE Transactions on Parallel and Distributed Systems, vol.11, no.6, pp.615-624, June 2000.
- J. Ho and F. Akiyildiz, âÃâ¬ÃËLocal Anchor Scheme for reducing Signaling Cost in Personal Communication NetworksâÃâ¬Ãâ¢, IEEE/ACM Transaction on Networking., vol. 4, no. 5, pp. 709-725 Oct. 1996.
- R. Jain and Y. B. Lin, âÃâ¬ÃËAn auxiliary user location strategy employing forwarding pointers to reduce network impact of PCSâÃâ¬Ãâ¢, ACMBakerJ. WirelessNetwork, vol. 1, no. 2, pp. 197-210, July 1995.
- W. Ma, Y. Fang, âÃâ¬ÃËTwo Level Pointer Forwarding Strategy for Location Management in PCS NetworksâÃâ¬Ãâ¢, IEEE/ACM Trans. On MobileComputing, vol. 1, no. 1, January -March 2002.
- K. Ratnam, I. Matta, and S. Rangarajan, âÃâ¬ÃËAnalysis of caching-based location management in personal communication networksâÃâ¬Ãâ¢, in IEEE ICNP '99, 1999.
- Yungoo Huh, Cheeha Kim, âÃâ¬ÃËNew Caching-based Location Management Scheme in Personal Communication SystemsâÃâ¬Ãâ¢, IEEE transaction, 2001.
- Jae-Woo LEE, âÃâ¬ÃËMobility Management Using Frequently Visited Location DatabaseâÃâ¬Ãâ¢, International Conference on Multimedia and Ubiquitous Engineering, IEEE transaction 2001.
- Jie Li, Yi Pan, Yang Xiao, âÃâ¬ÃËA Dynamic HLR Location Management Scheme for PCS NetworksâÃâ¬Ãâ¢, IEEE INFOCOM, 2004.
- Jae-Woo LEE, âÃâ¬ÃËA Model of Location Management Based on Broadcasting Visitor LocationâÃâ¬Ãâ¢, International Conference on Multimedia andUbiquitous Engineering, IEEE 2007
|