Multiplex communications – Data flow congestion prevention or control – Control of data admission to the network
Reexamination Certificate
1999-12-30
2003-12-09
Chin, Wellington (Department: 2664)
Multiplex communications
Data flow congestion prevention or control
Control of data admission to the network
C455S423000, C455S453000, C370S252000, C370S331000
Reexamination Certificate
active
06661776
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to estimating the call capacity in wireless communication systems.
2. Description of the Related Art
Recent developments in wireless communication systems have focused on increasing the capacity of wireless radio links. While these advances allow more calls to be processed at each base station, they may produce a situation where the processing capacity of a base station, rather than the radio link capacity, limits overall system performance.
In wireless communication systems, call control functions are typically performed by a single processor located in a base station. These call control functions tend to be transient events involved with call setup and handoff. For example, these call control events include processes executed by the call control processor to enable setting up of a call, handing off a call from one base station to another, and releasing a call. The steady-state traffic signals involved with a call-in-progress are not considered call control functions.
It is common to define “processor occupancy” as the fraction of time a processor is busy. For example, if the processor is busy performing operations for N seconds out of a T second period, the processor occupancy (expressed in percent) is (N/T)* 100.
The processor occupancy can be thought of as consisting of three components: no-load occupancy; call control processor occupancy and the occupancy due to operations; administration and maintenance (O, A & M). Of course, all processors have overhead associated with running the operating system, transferring data, etc. Thus, it is useful to define “no-load processor occupancy” as the processor occupancy that is not directly a result of call processing tasks. In other words, the no-load processor occupancy is the processor occupancy when no calls are being processed (i.e., under a so-called no-load condition). Similarly, “call control processor occupancy” is defined as the processor occupancy due to the above-described call control functions.
“O, A & M processor occupancy” is defined as the processor occupancy due to the O, A & M functions. Because of the real time needs of call control functions, the operating system in the processor is typically preemptive with static or dynamic priorities, meaning that certain events must wait until other, higher priority events execute. Processes that execute O, A & M functions are assigned priorities lower than those assigned to call control processes, because O, A & M tasks typically have loose delay requirements, meaning that these tasks may be delayed a relatively long time. Processes that are responsible for operating system specific functions and which contribute to no-load occupancy have extremely tight delay requirements and are assigned priorities higher than those assigned to call control processes. As a consequence of the preemptive nature of the operating system, O, A & M processes cannot execute whenever there is any operating system (i.e., no-load) or call control processes that are runnable.
Therefore, from the point of view of call control functions, the relevant occupancy components that contribute to call processing delays are those due to no-load and call control. Thus, in subsequent discussions, “total processor occupancy” will be understood to be the sum of no-load processor occupancy and call control processor occupancy. The call control processor occupancy may be calculated if the set of all call control events, their rates, and their respective processing times is determined.
Typical call control events which contribute to the processor occupancy of a base station will now be discussed. For the purposes of discussion and for ease of description, advanced mobile phone service (AMPS) and time division multiplex access (TDMA) call control events will be primarily described. However, the following call control events are generic and intended to be representative of like events in other systems, and of like events in different AMPS and TDMA implementations from different manufacturers. All of the following call control events contribute to the processor occupancy of a base station, each call control event having an associated processing time
Call events may be grouped into the following eight categories, for the purposes of description: mobile originations, mobile terminations, call release, handoffs out of a base station, handoffs into a base station, pages, registrations, and locates. These eight general categories will now be more specifically described.
The phrase mobile originations refers to calls originated by a mobile unit. This call event includes all of the messages involved in setting up a call. For example, this event typically includes messaging to and from the base station which detects a call attempt, messaging to and from the mobile switching center (MSC) which may determine the channel to be assigned and messaging to and from the mobile which results in the mobile tuning to the assigned channel.
The phrase mobile terminations refers to calls originated in the PSTN and terminating at a mobile unit. This call event includes all of the messages involved in establishing such a call.
Call release, the ending of a call, may be either mobile-initiated or forced by the MSC.
Handoffs out of a base station occur when a mobile served by a base station moves into an area better served by another base station. Handoffs into a base station occur when a mobile moves into the area better served by the base station. Both of these handoff events are included when calculating the processor occupancy.
Pages are messages transmitted from a base station to all mobile stations within its coverage area (i.e., cell).
Autonomous registrations for AMPS and TDMA systems occur when a mobile unit “announces” its presence to the nearest base station (and thus to the MSC). Once this presence has been registered with the base station's MSC (and with the mobile unit's home area MSC, if applicable), mobile terminated calls can find the particular mobile unit for which they are intended.
The final class of call control events are those used in the coordination of measurement activity prior to handoffs, generically referred to as locate events. Such locates are typically used to find the best cell, and/or antenna face within a cell, to serve an existing call. Locate events may be generated within the cell (Self Locate) or received from neighbor cells (Neighbor Locate Requests Received). Locate events occur in both AMPS and TDMA systems, though the manner in which they occur is different between these systems Having thus defined several classes of call control events, the processing time for each event may be measured for a particular processor, and for both AMPS and TDMA systems. The measurement times include only the actual processing time spent, and do not include elapsed time spent waiting for responses (either from the mobile unit or the MSC). These measurements may be single thread (single call) measurements made in an unloaded system using a measurement tool suited for such a purpose.
Call processing times may change with the addition of features in a particular system, and with each new software release for the system. Hence, measuring these processing times for each new release/feature may be made part of the standard laboratory testing program
In both AMPS and TDMA systems, when a specified call control event occurs a counter may be incremented. The number thus accumulated may be used to calculate the rates of the above-described call control events. Such counts are known in the wireless art, and detailed descriptions of these counts will vary between manufacturers of wireless systems. An example of call control event counts totaled on an hourly basis is shown in FIG.
1
.
Once the processing time of each call control event is known, and the count (i.e., # of a particular call control event per unit time, such as an hour) is measured, the processor occupancy as a percentage may be calculated for that particular call control e
Kaufman Joseph S.
Sampath Ashwin
Chin Wellington
Ha Yvonne Q.
Lucent Technologies - Inc.
LandOfFree
Method for estimating the processor occupancy and call... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for estimating the processor occupancy and call..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for estimating the processor occupancy and call... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3135522