Bandwidth Reservation Scheme in GSM
Current techniques for bandwidth reservation and allocation of handoff/new calls assume that the mobility pattern of the Mobile Station (MS) utilized profile-based schemes. This assumption may not be valid for practical systems. e.g. in wireless Asynchronous Transfer Mode (ATM) network environments, wireless components can be connected to Wide Area Network (WAN), Local Area Network (LAN) or even Home depending on the type of ATM network is to be accessed. For such varied wired networks, it may not be possible to predict the arrival of MS to some cell since the mobility patterns may not be available. And varying traffic conditions suggest that such history-based schemes can never be fully reliable. So in this scheme the real time position of MS will be measure to calculate or find the next path of moving MS. The very most advantage of real time positioning is to determine the next cell that the MS will cross with high accuracy. Therefore the reservation of wireless resources is only done the next cell among all of the neighboring cells. In this way it eliminates the reservation of excessive bandwidth in neighboring cells. This is done with consideration to the limited radio resources as compared to the wired component of wireless network.
1.1 Bandwidth Access Approach:
Typically, a resource in cellular networks refers to the available bandwidth (such as time slots in the Time Division Multiple Access (TDMA) scheme, codes in the Code Division Multiple Access (CDMA) scheme, or frequency bands in the Frequency Division Multiple Access (FDMA) scheme), transmission power level, and the amount of buffer allocated for accommodating the incoming calls in the BS. For simplicity, here the only bandwidth will be considered. To deal with the multiple classes (multimedia traffic) of calls; the calls may differ in traffic characteristics (Constant Bit Rate--CBR, Variable Bit Rate--VBR and Unspecified Bit Rate--UBR) and the desired Quality of Service (QoS) guarantees, i.e. delay bound, loss bound and throughput. The concept of effective bandwidth in [10] is adopted from a wired network instead of a wireless network, where effective bandwidth means the minimum amount of bandwidth needed to provide a specific QoS given the traffic parameters of a call and the buffer size at the multiplexer. Due to distinct characteristics of wireless cellular networks, it is necessary to develop mechanisms tailored to support QoS for MSs who run multimedia services. QoS in wireless cellular networks can be considered and achieved at two abstraction levels:
a. Call-level QoS: related to connection establishment and management, which are very important especially in dealing with user mobility. In wireless cellular networks, a user’s QoS requirements can be quantitatively expressed in terms of probabilistic call-level QoS parameters such as New Call Blocking Probability (NCBP) and Handoff Call Dropping Probability (HCDP).
b. Packet-level QoS: a MS at this level is concern with metrics such as like maximum packet delay, maximum jitter and packet dropping probability. MSs need guarantees on one or more of these quantities so that the receiver can recover meaningful information from the transmitted message. Packet scheduling methods are responsible for distributing the radio network resources among the MSs within the coverage area. A packet scheduling method is utilized at the packet level.
There is assumption that the system will allocate a certain number of bandwidth units to different classes of calls. In this scheme, a one-to-one matching scheme is proposed. In other words, a Bandwidth Block (BB) can only be used by the MS that reserved it. The term BB comes from the fact that in a normal case a handoff call belonging to a certain class will use a series of allocated time slots. There are two reasons for doing it as follows:
C. In CDMA cellular networks, a MS can communicate with multiple base stations at the same time when it is located inside a so called boundary region of multiple cells. It adaptively chooses the BS with the best communication quality. In this case, it not only holds a channel in its primary cell, but also holds (or reserves) channels in other adjacent cells. This approach used in CDMA cellular networks for Soft handoff.
D. Assuming MS1 reserved a time slot in some carrier frequency, which is a common case in mobile multimedia that use TDMA as media access protocol and it has a highest priority (serve MS1 beforeMS2) while MS2 reserved five time slots. If we use a shared approach, there should be no differences among reserved time slots. It is possible thatMS1 is assigned time slot No. 1 and MS2 is assigned time slot No. 2–6. Thus MS2 cannot obtain contiguous time slots such as No. 1–5. This is generally not accepted in a multi-rate system. It can also bring difficulties for the implementation of actual signal sampling hardware components.