CIT'04 Proceedings of the 7th international conference on Intelligent Information Technology
| Hi-index | 0.00 |
Distance computation is the key to the design of several adaptive location management schemes (distance-based, speed-based etc.) for cellular mobile networks. Distance can be measured in terms of number of cells Mobile Unit (MU) is far from the last updated cell (LUC). Distance computation is a real challenge as updates are triggered by the mobile unit (MU), which does not have knowledge of the service area layout. A novel technique using Cell Identification Code (CIC-DIS) exists in literature, which encodes each cell with a 4-bit locally unique code. Considering hexagonal cells arranged in a 2-D rectangular grid, the cell in the ith row and jth column would have the code [i%3 j%3] in binary. CIC-DIS coding works only for uniform 2-D grid of hexagonal cells. We propose an efficient direction based cell identification coding (CIC-DIR) technique, which an enhancement of the CIC-DIS. In CIC-DIR each BS continuously broadcast through the downlink control channel in GSM a sequence of 4-bit codes according to the following rule:1st code contains the CIC of the cell to which the BS belongs. Next six codes contain CIC of its six neighbors in a certain order that is to be followed by all BSs. The 8th code in the sequence is a special code 1111, used only to delimit the sequence (should not be used as CIC). Each MU is required to store only the current cell CIC (c_cic) in memory. Whenever the MU moves to a new cell it first synchronizes with respect to the new sequence by identifying the delimiter code 1111, then checks the relative position of c_cic within the new sequence, say x. This x is identified as the side or direction with respect the last cell through which it moves. To compute distance, the cells surrounding the LUC are considered to be organized in several concentric rings with the LUC at the center. Therefore, it is obvious that the displacement value will be same as the ring number to which the new cell belongs. It is observed that the total set of cells Ctot in each ring can be divided in two subsets C222 and C321. The six sides in each cell can be divided in three subsets S+, S−− and S0 transition through which causes distance value to be incremented by one, decremented by one and remain unchanged respectively. Cells belonging to subset C222 will have two side members each in S+, S−−and S0, whereas cells belonging to subset C321 will have three side members in S+, one in S−− and two in S0. It is observed that the Nth cell in Rth ring belongs to C321 if ((N%R)==0). After each move the new ring number R’(distance value) will be equal to R + 1, R – 1or R depending on whether side x through which it moved belongs to S+, S−− and S0 respectively, conditions for which are derived. The new cell number N’ can also be calculated from N and R easily after each move. The MU requires storing values of c_cic, N and R only. Thus the distance computation is simple and can be implemented with little additional overhead.