Impact of mobility in mobile broadband systems multi-service traffic

Abstract

Multi-service traffic engineering will be a key aspect in cellular planning of Mobile Broadband Systems (MBS), allowing to obtain merit functions for optimisation purposes. MBS will serve applications via different service components, with different data rates and average durations, user mobility having impact on the handover failure probability threshold. While in the business city centre and other urban scenarios mobility has no significant effect, it affects the supported traffic in main roads; a reduction of 31 % existing in the 432 channel/cell case, relatively to the situation of absence of mobility.

Mobile World

“…There are many more specialist Data Applications designed for particular industries, including mobile surveillance, urban guidance, tourist guides, parcel tracking, etc. (Velez and Correia, 2001). These Data Applications generate particular demands on service infrastructure, utilising one or more types of network services.…”

High capacity wideband traffic in enhanced UMTS: a step towards 4G

“…Although they enable to capture some degree of correlation of traffic, they present an inadequate autocorrelation, and are unsuitable for LRD modelling. Hence, MMPP are only adequate to represent the average behaviour of the aggregate traffic, e.g., for cellular planning purposes [11], but not for dimensioning traffic management functions or assessing detailed QoS parameters. Since the traditional traffic models are becoming inadequate to capture today's network characteristics, traffic models for packet-switched data have been developed on the basis of measurements from actual data networks.…”

“…The probability of forced termination of a connection during its duration, i.e., the connection dropping probability, P d , can be associated to the latter [5]. Given QoS constraints for blocking and connection dropping probabilities (P b = 2%, and P d = 0.5%), the BPP model is used to obtain the supported load; details on the model itself, the service components, and the user model can be found in [11]. In E-UMTS one can consider that resources/channels serve applications via different service components, Table 1, i.e., the system itself serves service components, which, in turn, serve applications.…”

“…Different types of mobility are assumed for each application in each of the scenarios. A triangular distribution is considered for the velocity, with average V av and deviation ∆ [8,11]. Values for V av = ∆ = 0, 1, 10, 15 m⋅s -1 are considered for the ST, PD, UB, and MR scenarios, respectively, while V av = 22.5 and ∆ = 12.5 m⋅s -1 for the HW scenario.…”

“…It is often discussed in literature how Moore's and Gilder's laws can be generalised to cellular communication systems. However, it is not enough to say that typical data rates will increase, e.g., from 9.6 to 64 kb/s in GPRS, to 384 kb/s within UMTS, or to 2 Mb/s within E-UMTS, or even 10 Mb/s within MBS (Mobile Broadband Systems) [11]. It is also necessary to compare system capacity, e.g., in kbps/km, and to verify what its actual evolution will be.…”