Monday, March 18, 2013

• Capacity Calculations

The capacity of a GSM-based network is dependent on the frequency reuse scheme and the amount of the operator’s spectrum.  Frequency reuse schemes are used to divide the spectrum into unique groups of channels, so that the frequency planner can methodically assign channels to a network of cells.  The reuse scheme is chosen to minimize the amount of interference in a network of cells, but also has an impact on the capacity of a particular site. 


The amount of spectrum a license holder has must be considered as well.  Since the GSM channel is 200 kHz wide, five channels can fit into 1 MHz of spectrum.  Therefore, a European operator, with 25 MHz of spectrum, will have 62 channels available;wherea, a PCS operator in the United States, with 30 MHz of spectrum, will have 75 channels available.  Remember that half of the spectrum is for downlink and half is for uplink.


A GSM market that has just launched commercial service will likely use a 4/12-reuse scheme.  Using a 4/12-reuse scheme will have less capacity (not likely to be a problem as the network turns on), but more importantly, the larger distance to reuse will minimize interference.  As subscribers are added to the network, the capacity increases to the point that frequencies must be reused over shorter distances.  Using frequency hopping, the GSM network may be frequency planned using all of the channels within three sites.  Frequency hopping is explained in a later section.

The number of channels available per sector when considering the two reuse schemes is calculated as follows:
To calculate the capacity of the network, the Erlang capacity of each sector must be identified.  The Erlang capacity of each sector is calculated by  considering the desired Grade Of Service (GOS) and the number of voice channels per sector.  It is industry-accepted practice to design a network with 1 - 2% GOS.  The number of voice channels per sector is calculated with respect to the number of RF carriers per sector.  Generally, eight time slots are considered for each carrier except the first carrier, which dedicates one or two time slots for control and messaging.  By using an Erlang-B table, the Erlangs per sector are determined.

The Erlang capacity of the network is calculated as follows: 
Comparing the total capacity of networks using the two different reuse schemes is depicted  as follow.

                                Capacity of Network Using 3/9 Reuse Scheme
Total Number of Cells Required
100

Number of Voice Channels/ Sector
22

Blocking Rate
1
%
Erlangs/Sector
13.65

Sectors per Site
3

Erlangs/Site
40.95

Erlang Capacity of Network
4095


                                   Capacity of Network Using 4/12 Reuse Scheme
Total Number of Cells Required
100

Number of Voice Channels/ Sector
14

Blocking Rate
1
%
Erlangs/Sector
7.351681

Sectors per Site
3

Erlangs/Site
22.05

Erlang Capacity of Network
2205


In the case of a network using a 3/9-reuse scheme, there would be 810 channels.  Conversely, the same network of sites using a 4/12-reuse scheme would have 630 channels; however, the capacity has almost doubled.  This example is a bit over simplified, and conservatively assumes that all sites in the network have the same distribution of traffic; however, it depicts the gains associated with a tighter reuse scheme.

You might also like:




No comments:

Post a Comment