Telecommunications – Radiotelephone system – Zoned or cellular telephone system
Reexamination Certificate
1998-10-13
2002-05-07
Kincaid, Lester G. (Department: 2682)
Telecommunications
Radiotelephone system
Zoned or cellular telephone system
C455S440000, C455S452200, C455S456500
Reexamination Certificate
active
06385454
ABSTRACT:
TECHNICAL FIELD
The invention relates to the management of connections to mobile units in wireless communications systems and, in particular, wireless cellular communications systems.
BACKGROUND OF THE INVENTION
Wireless cellular communication systems require efficient mobility management algorithms to cope with frequent handoffs and rerouting of connections, as the mobile units frequently change their points of attachment to the network. Consequently, one of the key challenges for the designers and managers of such systems is to develop a system, which adapts to the mobile unit's mobility by managing connections efficiently in a manner that ensures that an acceptable quality of service (QoS) is provided to the mobile unit.
Mobility management entails both connection management and location management. Connection Management contains both a connection-establishment phase prior to data exchange and a connection-release phase after data exchange. In a wireless network, as mobile units move through cells of the network, connections have to be torn down and re-established with a frequency that corresponds to the speed of the mobile through the cells. Managing the connection during a transition from one cell to another while maintaining the integrity of the data, includes preserving the packet sequencing, preventing packet duplication and avoiding loss of packets.
At the most fundamental level of QoS requirements, is the ability of the network to maintain connectivity with the mobile unit even when the terminal frequently changes its physical location. It is possible to maintain connectivity with the mobile unit and guarantee a certain QoS to it if the network knows prior to the mobile's movement the exact trajectory it will follow. With this information, the network can determine if there are enough resources available along the mobile's path for the lifetime of the connection. If such is the case, the network anticipates the mobile's demands and takes appropriate steps such as setting up end-to-end routes, reserving resources along these routes and planning quick low-latency handoffs between base stations of adjacent cells. With these kinds of preparations QoS can be guaranteed.
An approach for providing QoS to mobile terminals was recently proposed by A. Acampora and M. Naghshineh, in their paper “An Architecture and Methodology for Mobile-Executed Hand-off in Cellular ATM Networks,” published in the
IEEE Journal on Selected Areas in Communications,
Vol. 12, No. 4, October 1994, pp. 1365-1375. In this paper the authors propose a technique called the virtual connection tree (VCT) scheme in which connections are pre-established between a fixed (root) switch and a set of base stations with whom the mobile could potentially connect. The VCT approach maintains QoS by pre-establishing end-to-end connections from the base stations with which the mobile unit could potentially attach, and consequently by minimizing handoff latency between base stations (i.e., the time between initiation and completion of the hand off) and by minimizing packet loss, which interrupts the connection. Unfortunately, the VCT approach results in inefficient use of network resources, is a potential source for overloading the network, and requires substantial processing time for setting up and assigning the Virtual Connections (VC). Network inefficiencies occur since the VCs are pre-assigned without accurately taking into account the mobile's current and projected movement patterns. Consequently, many pre-assigned VCs are wasted and efficient resource reservation cannot be achieved.
The VCT approach suffers from the lack of accurate knowledge of the mobile's trajectory. As a result, there is a substantial risk that the connection resources at the base stations will be under utilized, with the mobile never connecting to a reserved channel and potentially overloading the base stations with large numbers of unused reserved channels. Overloading can lead to traffic congestion in a cell, which may result in the base stations either dropping or buffering a connection. Buffering can cause temporary violations of the network's delay and cell loss guarantees and can effect QoS.
As a possible solution to these kinds of problems caused by the VCT approach, the concept of a Shadow Cluster has been proposed. A Shadow Cluster defines the area of influence of a mobile terminal (i.e. a set of base stations or network cells to which the mobile unit is likely to attach in the near future). Like a shadow, this set moves along with the mobile, incorporating new base stations while leaving the old ones as they come under and out of the mobile's influence. Each base station in the Shadow Cluster anticipates the mobile's arrival and reserves resources for it. A close association exists between the predicted time the mobile will arrive at one of the cells of the Shadow Cluster and the time when the cell's resources are reserved. The accuracy of the prediction of the mobile's path determines the number of base stations in which resources are actually reserved and consequently determines the overall network overhead required to maintain the desired QoS.
Location management or location tracking incorporates a set of mechanisms with which the network can locate a particular mobile at any given time. Location updating and location prediction are two strategies that can possibly be used to implement mechanisms for locating a mobile. Location updating is a passive strategy in which the network periodically records the current location of the mobile in a database. Thus, the proficiency of the mobile tracking algorithm depends on the frequency with which the location of the mobile is updated, which in most network systems is controlled by the mobile. In contrast, location prediction is a dynamic strategy in which the network system estimates the mobile's location based on a model of the mobile's movement. The proficiency of the tracking depends on the accuracy of the model and the accuracy of the algorithm used to predict the future movement of the mobile.
While most recent studies have focused on the updating method, relatively little has been done with respect to the prediction approach. As a consequence, management or tracking of a mobile is generally treated as purely a process of updating and querying databases. If accurate prediction of the movement of a mobile was possible, the task of locating mobiles given their last location would become substantially more efficient in terms of both speed and use of system resources.
One way for the network to know the future direction of the mobile is to have a formal mechanism in place that allows the mobile user to indicate to the system his or her intended destination and the duration of the connection. The network can then combine this information with its knowledge of the geography of the terrain and the location of the base stations within the terrain to determine the cellular path of the mobile. Unfortunately, this is not an all-encompassing solution since there may be multiple paths to a destination. Even after using the general direction information provided by the mobile the system cannot exactly determine which one of the multiple paths the mobile will follow on every occasion. It is reasonable to expect that the mobile may diverge from a system selected route without warning in order to adjust to a dynamically changing environment around it. Without the network also dynamically adapting to such a change in the expected cellular route, the amount of resources required to provide improved connectivity is prohibitively great and consequently unattractive and possibly even impractical.
Some previous works in the area of mobility trajectory prediction include a suggestion that the mobile's location be determined based on a behavior model represented as a set of historical movement patterns stored in a user profile. This model can be further refined to be modeled as repetitions of some elementary historical movement pattern
Bahl Paramvir
Liu Tong
Kincaid Lester G.
Leydig , Voit & Mayer, Ltd.
Microsoft Corporation
LandOfFree
Apparatus and method for management of resources in cellular... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Apparatus and method for management of resources in cellular..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Apparatus and method for management of resources in cellular... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2861966