12.3 Radio Resource Management Procedures

12.3 Radio Resource Management Procedures

12.3.1 Power Control

The support of power control procedure is mandatory in the uplink. The procedure includes both an initial calibration process and a periodic update process.

By measuring the received power at the BS side, the BS can send power offset indications to the MS, which adjusts its transmit power level accordingly. It has to be noted that in the case where the MS uses only a portion of the subchannels, the power density remains unchanged regardless of the number of subchannels actually used (unless the maximum power level is reached).

The power offset indication sent by the BS is consistent with the actual MCS that is to be used for the uplink transmission. Power adjustments related to the use of a different modulation scheme is then taken into account. Finally, the power offset is also consistent with the maximum power that the MS can transmit (indicated in the SBC-REQ MAC message).

After reception of the power offset information, the MS adjusts its transmit power according to Equation (12.1) by simply adding the offset to the last value used for transmission:

where P_last is the last used transmit power, P_new the new transmit power and ‘Offset’ the accumulation of offset values sent by the BS since the last transmission. Transmission power offset information is sent in RNG-RSP messages in units of 0.25 dB. This process allows for power fluctuations at a rate of at most 30dB/s with a depth of at least 10dB.

In the case of OFDMA-based WiMAX systems, the power control differs for some uplink burst types. For uplink burst regions used for the fast feedback channel (UIUC = 0), CDMA ranging (UIUC = 12) and CDMA allocation IE (UIUC = 14), the transmit power update formula of the MS is the following:

where

• P_New = P_Last and ‘Offset’ are as defined in Equation (12.1),

• C/N_new is the normalised C/N (Carrier over Noise) of the new MCS used in the region,

• C/N_last is the normalised C/N of the last used MCS,

• R_new is the number of repetitions of the new MCS used in the region,

• R_last is the number of repetitions of the last MCS.

The normalised C/N values are default values given by the standard (see Table 12.1). However, the BS may override these values using a dedicated UCD message TLV.

Table 12.1: Normalised C/N values for power control procedures for OFDMA-based WiMAX terminals. (From IEEE Std 802.16e–2005 ^[2]. Copyright IEEE 2006, IEEE. All rights reserved.)
Open table as spreadsheet

MCS	Normalised C/N(db)
ACK region	−3
Fast feedback region	0
CDMA code	3
QPSK 1/3	0.5
QPSK 1/2	6
QPSK 2/3	7.5
QPSK 3/4	9
16-QAM 1/2	12
16-QAM 2/3	14.5
16-QAM 3/4	15
16-QAM 5/6	17.5
16-QAM 1/2	18
16-QAM 2/3	20
16-QAM 3/4	21
16-QAM 5/6	23

The power control mechanism may also be implemented in the downlink (by, for example, limiting the interference created to the other cells). However, its implementation, if any, is vendor-specific.

12.3.2 Dynamic Frequency Selection (DFS)

The DFS mechanisms may be required in the case of deployment of a WiMAX system in a license-exempt band (e.g. the 5.8 GHz band). In that case, the BS and the MS implement a set of mechanisms that permits:

• sounding the radio environment prior to the use of a channel;

• periodically detecting ‘specific spectrum users’ (a specific spectrum user is a user that has been identified by the regulator as requiring strict protection from harmful interference);

• discontinuing operation on a channel after detection of a specific spectrum user;

• scheduling of periodic sounding testing periods (from the BS and the MS),

• selecting/changing to a new channel.

In any case, the BS cannot use a channel without testing the channel for other users, including specific spectrum users.

An example of a simplified process flow for DFS operation is depicted in Figure 12.1. At initialisation, the BS sounds the channels based on predefined timing parameters. Before potentially using a new channel, the BS must sound this channel for at least a ‘startup test period’ against other users. After completing the scanning of the channels, the BS can choose to operate on a channel that is not used by specific spectrum users and can then establish a connection with the MS in its coverage area. The mechanisms to sound the channels, to measure the interference, to detect specific spectrum users and to select a channel for operation are left to the manufacturer.

Figure 12.1: Example of a simplified process flowchart for a WiMAX system implementing DFS

During the operation on a channel, the BS may be assisted by the MS to measure the use of one or more channels by other users. The scheduling of this process can be done either during a quiet period in the cell or during normal operation. When the BS requests MS support, the BS informs the MS by a channel measurement IE in the DL-MAP. Then, if the channel to be measured is the operating channel, the BS suspends all transmission/scheduling of the MAC PDU to any MS in the cell area during the measurement interval. During the testing process, the MS stores several parameters: the frame number corresponding to the first measurement, the accumulated time of measurement and the existence of a specific spectrum user in the channel. Those parameters are reported back to the BS during a measurement report response. However, if the MS detects a specific spectrum user, it will send an unsolicited REP-RSP message to the BS as soon as possible.

Upon detection of a specific spectrum user on the operating channel (either by the BS or the MS), the BS must discontinue the transmission of data MAC PDUs and MAC management message MAC PDUs within predefined periods (‘Max Data Operation Period’ and ‘Management Operation Period’ respectively).

When the operation needs to be discontinued, the BS starts to select a new channel for operation either from a recent and valid tested channel set or by resuming a similar process to that used at initialisation. After the selection of a new channel, the BS informs its associated MSs about channel change by including in the DCD message the new channel number and the frame number where the switch occurs.

In the case of DFS, the regulatory authority defines the timers involved in the sounding procedure channel (during initial sounding or periodic sounding) as well as the thresholds in order to prevent harmful interference to other users.

Finally, a similar mechanism can be implemented for other purposes than DFS and may be applied to any WiMAX deployments in shared channel environments.

12.3.3 Other Radio Resource Management Procedures

To optimise the performance of IEEE 802.16-based systems, other radio resource management procedures are implemented. The admission control of new connections is part of the RRM operation. The WiMAX Forum defines a framework to support and optimise the admission control (see Section 12.3.5). Admission control decision algorithms are left to the manufacturers.

Link adaptation mechanisms are also implemented. Again, the way the process operates and the selection of the MCS according to channel conditions and other local criteria are left to the vendor. More details on the supporting primitives for link adaptation are provided in Chapter 11.

12.3.4 Channel Measurements

In order for the BS to take appropriate decisions for radio resource management (power control, selection of the modulation and coding scheme and use of advanced antenna technology), the standard defines a set of channel quality indicators. Two families of indicators are available:

• RSSI (Received Signal Strength Indicator), which gives information on the received power level;

• CINR (Carrier-to-Interference-and-Noise Ratio), which gives information on received carrier-to-interference levels.

WiMAX radio equipment (MS and BS) can implement the means to measure, compute and report these indicators. In addition, to reflect the fluctuations of the radio channel in time, two statistics of the indicators are evaluated and reported: the mean and standard deviation.

12.3.4.1 Received Signal Strength Indicator (RSSI)

RSSIs are derived from measured received power level samples. The reported RSSIs are an average (in linear scale, e.g. in mW) of the measured power level samples (averaging is done by an exponential filter with a forgetting factor provided as a configuration parameter by the BS).

The mean RSSI is obtained from

where R[k] is the measured power sample during message k and α_avg is the averaging factor. The sample index k is incremented for every frame and the power is measured over the frame preamble. The averaging factor is transmitted by the BS either in a DCD message TLV or in a REP-REQ MAC message. Otherwise, the default value of 1/4 should be considered by the MS.

The standard deviation of RSSI is derived from the expectation-square statistics of the measured signal levels, x²_RSSI, defined by

The method to measure the received signal strength is vendor-specific. However, the measurements should remain inside a ± 4 dB absolute error.

The reported values are sent in MAC REP-REQ message using the dBm (dB) scale for the mean RSSI (respectively the standard deviation), as specified by

The values are quantized with a 1 dB increment and each is sent in a 1-byte field in the MAC REP-REQ message. The range of RSSI spans from −123 dBm to −40 dBm.

12.3.4.2 Carrier-to-Interference and Noise Ratio (CINR)

For the IEEE 802.16-2004-based systems (using OFDM), the reported CINR indicators are physical CINR indicators. When the BS requests a CINR measurement report from the MS. the MS will answer back by including in the REP-RSP MAC message the estimates of the average and standard deviations of the CINR. CINR values are reported in a dB scale with 1 dB steps, ranging from −10 dB to 53 dB.

Reported CINR values are averaged using the same averaging method as that for the RSSI. The method to evaluate the CINR is vendor-specific. The measurement samples can be taken either from detected or pilot samples.

For the IEEE 802.16–2005-based systems (using OFDMA), additional options are specified. On the one hand, different CINR measurements are defined: physical and effective CINR measurements. On the other hand, the CINR reports may be sent either through the REP-REQ MAC message (mean and/or standard deviation of CINR) or through the fast-feedback channel (mean CINR only).

Several physical CINR reports can be requested from the MS:

• Physical CINR measurements on the preamble. In the case of frequency reuse 3 networks, the CINR is measured over modulated carriers of the preamble, while in the case of the frequency reuse 1 network, the CINR are measured over all the subcarriers (modulated or not, excluding the guard bands and DC channel).

• Physical CINR measurement from a permutation zone. In this case, CINR samples are measured from the pilots in the permutation zone.

In addition, the BS may also request ‘effective CINR’ measurements from a permutation zone on the pilot subcarriers. The effective CINR is a function of the physical CINR, taking into account channel conditions and implementation margins (implementation is manufacturer-dependent).

With this option, the MS has the additional possibility of indicating the MCS that best fits the specified target error rate to the BS.

12.3.5 Support of Radio Resource Management in the WiMAX RAN

12.3.5.1 RRM Functional Spit in the WiMAX RAN

The WiMAX Forum also defines a framework to optimise the operation of RRM in a WiMAX radio access network (also called an Access Service Network (ASN), see Chapter 13), and also supports multivendor interoperability of RRM procedures in the long term.

As specified by the WiMAX Forum, the RRM function in the network may provide reporting facilities and decision support to several network functions, such as:

• admission control, e.g. ensuring that enough radio resource is available at the BS side to serve appropriately a new MS or connection or service flow (either at service request or after a handover);

• handover preparation and control, e.g. optimising the choice of the target BS according to radio and BS load indicators.

Optionally, the RRM may also be involved in transport network resource management.

This framework is based on a functional split of the RRM functions into two parts (Radio Resource Agent (RRA) and Radio Resource Controller (RRC) that communicate through standardised primitives ^[21].

The primitives exchanged are used either to report information (from RRA to RRC or between RRCs) or to communication decision support information (from RRC to RRA). This information includes measurement reports per MS and spare capacity per BS.

12.3.5.2 Future Enhancements of the RRM Network Function

As of today, the BS autonomously and independently performs the power control and interference management procedures. However, the RRM framework defined by the WiMAX Forum leaves the door open for further enhancement of the RRM procedures. Possible enhancement could be done by exchanging additional information (e.g. on channel configurations) between RRM entities in order to have a global optimisation of radio resource and not only a BS per BS optimisation.

^[21]WiMAX Forum Document, WiMAX end-to-end network systems architecture; Stage 2, Release 1: architecture tenets, network reference architecture, reference points, April 2006.