WiMAX solutions include fixed WiMAX and mobile WiMAX. According to the type of terminal offered (outdoor CPE (Consumer Premise Equipment), self-install indoor CPE, mobile terminal–PCMCIA, laptop, smart phones), each solution has a different deployment method.
Coverage from a BS is linked to several radio parameters. The first one is the propagation environment. Depending on the relative locations of the BS and the terminal, several models can be used to evaluate the losses due to propagation.
Line-of-Sight (LoS) and Near Line-of-Sight (NLoS) propagations may happen when the BS and the MS are deployed outdoor, above the average height of the environment. This may be the case of the deployment of a fixed WiMAX solution in a rural environment with the BS located on a high altitude point (or at the top of a mast) and the SS (then a CPE, for a fixed WiMAX) deployed on the rooftop of the customer's house.
LOS propagation requires that the first Fresnel zone is free from any obstacle. In that case, the propagation losses are proportional to the square of the distance between the BS and the SS . However, in a practical deployment, this happens very seldom. Usually, the Fresnel zone is obstructed and/or there are few obstacles on the BS to the CPE transmission path. In that case, near LOS models are used. However, in the case where there are few obstacles (a building, a water tower, a hill), the obstacles are modelled by knife-edge. An example of such a model can be found in Reference .
NLOS (Non-Line-of-Sight) propagation occurs when the terminal is located indoor and/or at ground level. In this situation, there is in most cases no direct path between the BS and the terminal, there is a high number of obstacles on the BS to MS path (buildings, trees, cars, etc.) and the receiver may receive several copies of signal that experienced several reflections/diffractions on different obstacles. This type of propagation is typical of cellular deployments.
In the case of WiMAX, NLOS propagation corresponds to a deployment of a fixed or mobile WiMAX using self-install indoor terminals (wireless DSL deployment), or a deployment of a mobile WiMAX with mobile terminals (PCMCIA, laptop with integrated chipset, multimode mobile phones, etc.).
In NLOS, the losses versus the distance are much higher that in LOS. Usually, the distance decay exponent is between 3 and 4. An example of a propagation model that can be used to evaluate propagation losses in NLoS is the Erceg model .
In order to evaluate the range, additional parameters are required. First, the value of some propagation parameters depends on additional radio parameters such as:
Frequency band. WiMAX could be deployed at different frequencies (2.3 GHz, 2.5 GHz and 3.5 GHz); the higher the frequency, the higher the propagation losses.
BS antenna height. Propagation losses decrease if the antenna height is increased.
Terminal antenna height. The lower the terminal, the higher the losses (for mobile deployment, the height of the terminal is usually taken to be between 1.5 and 2m).
Then, depending on the deployment scenario, a margin modelling the fluctuations of the radio environment needs to be considered.
In the case of the LOS/near LOS scenario, the BS and the MS are fixed. However, the propagation loss fluctuates (according, for example, to geoclimatic parameters). Hence, to evaluate the range, a margin needs to be evaluated according to the signal availability required (e.g. 99%) and the geoclimatic parameters. A method to evaluate the margin is available in Reference .
In the case of the NLOS scenario, in order to reflect the distribution of the obstacles in the coverage area, a margin, called the shadowing margin, needs to be taken into account for the evaluation of the range. This shadowing effect is modelled by a lognormal distribution with a standard deviation, which depends on the environment (typically from 10dB in an urban environment to 5 dB in rural environments). The resulting margin also depends on the signal probability availability on the cell area (usually between 90 and 95%). A method to evaluate this margin can be found in Reference . In addition, other margins may be included in the propagation losses: indoor margin (in the case of indoor coverage requirements), interference margin (to reflect the interference level generated by other cells transmitting at the same frequency), body losses, etc.
Finally, to evaluate the range, the system gain may be evaluated. The system gain is the maximum signal level difference that can be accepted on the BS-to-terminal path so that the receiver may decode a signal with sufficient quality. The system gains depend on many specific parameters:
Transmission power of the BS/terminal.
Antenna gain of the BS/terminal. The gain at the terminal side depends on the type of terminal. Outdoor CPEs may have gains in excess of 14dBi while mobile terminals may have only a 0dBi antenna gain.
Receiver sensitivity. The receiver sensitivity depends on the signal bandwidth, the modulation and the coding scheme. The IEEE 802.16 standards are providing minimum reference values but vendors may come with better performance solutions.
For fixed WiMAX solutions deploying outdoor CPEs, the coverage may be of several km (see Reference ). However, to get range figures in line with an exiting cellular system for WiMAX cellular-like deployments, additional features are needed to compensate for the extra losses due to lower terminal antenna gains, indoor penetration and NLOS behaviour.
The radio enhancement feature applicable to fixed and mobile WiMAXs is subchannelisation. This is the possibility to concentrate the transmit power on a few subchannels and thus to play with trade-off between the coverage and the maximum data rate a terminal can get at the cell edge. Other enhancement features that are only applicable to the mobile WiMAX are:
Convolutional Turbo Coding (CTC).
Repetition Coding. For each retransmission, up to 3 dB gain can be obtained. This again allows a trade-off between the maximum range and data rate at the cell edge.
Hybrid ARQ (HARQ). This is the capability of the receiver to combine several transmissions of the same MAC PDUs.
Advanced antenna technologies: beamforming and MIMO.
Again, according to the deployment scenario, the frequency planning guidelines are different.
In the case of a fixed WiMAX with an outdoor CPE deployed in LOS/near LOS, similar frequency planning to that of Fixed Wireless Access (FWA) systems (e.g. PMP microwave transmission) can be used. At the radio site, several alternatives for sector configurations may be employed (1, 3, 4 and 6 sectors per site). In FWA-like deployment, the CPE has an external antenna with high gain and hence a reduced antenna beam width. Consequently, the antenna of the CPE provides significant interference protection; hence in order to achieve the CINR target on the service area, frequency planning with 3 and 4 frequencies may be used. For the fixed WiMAX with an indoor self-install CPE, which also have directive antennas, a similar frequency plan may be used.
For densification, more sectors (up to six) or more frequencies per sector can be deployed on the existing sides.
In the case of a mobile WiMAX deployment with mobile terminals, there is no interference reduction due to the MS antenna (omnidirectional antennas). However, because of the use of OFDMA and permutations (which gives some randomness in the use of subcarriers), and because of the better radio performance (more coding capabilities, beamforming and/or MIMO), it is possible to deploy an efficient system with a frequency planning reuse scheme of 3 and even 1 .
In the case of MBS (see below), the frequency channel assigned for transmitting multicast/broadcast connections must be deployed in a Single Frequency Network (SFN) manner.
For densification, the same techniques used for existing cellular systems are employed: more frequencies per site, cell splitting, microcellular layer, indoor coverage from the indoor, etc.
IEEE 802.16-2004 indicates that there are three options of BS synchronisation :
Asynchronous configuration. Every BS uses its own permutation. The frame lengths and starting times are not synchronised among the base stations. This configuration can be used as an independent low-cost hot-spot deployment.
Synchronous configuration. All the BSs use the same reference clock (e.g. by using GPS, or Global Positioning System). The frame durations and starting times are also synchronised among the BSs but each BS may use different permutations. Due to the time synchronisation in this scenario and the long symbol duration of the OFDMA symbol, fast handovers as well as soft handovers are possible. This configuration can be used as an independent BS deployment with a controlled interference level.
Coordinated synchronous configuration. All the BSs work in the synchronous mode and use the same permutations. An upper layer is responsible for the handling of subchannel allocations within the sectors of the base station, making sure that better handling of the bandwidth is achieved and enabling the system to handle and balance loads between the sectors and within the system.
The standard indicates that, for TDD and FDD realisations, it is recommended (but not required) that all BSs should be time-synchronised to a common timing signal.
Recommendation ITU-R P.526–8, Propagation by diffraction, 2003.
Erceg, V., et al., Channel models for fixed wireless applications, IEEE 802.16 Broadband Wireless Access working group, February 2001.
Recommendation ITU-R P.530–9, Propagation data and prediction methods required for the design of terrestrial line-of-sight systems, 2001.
IEEE 802.16e, IEEE Standard for Local and Metropolitan Area Networks, Air Interface for Fixed Broadband Wireless Access Systems, Amendment 2: Physical and Medium Access Control Layers for Combined Fixed and Mobile Operation in Licensed Bands and Corrigendum 1, February 2006 (Approved: 7 December 2005).
WiMAX Forum White Paper, WiMAX deployment considerations for fixed wireless access in the 2.5 GHz and 3.5 GHz licensed bands, June 05.
WiMAX Forum White Paper, Mobile WiMAX - Part I: a technical overview and performance evaluation, March 2006.