MAC Layer Features

Quality of Service (QoS)

At a high level, QoS is achieved with the following features:

  • Service Class Definitions with QoS profiles that determine the parameters such as scheduling type, traffic rate, and latency
  • Mapping of higher layer traffic to Service Flows, which are instances of service classes
  • Per-connection queuing and buffering

Scheduling Service

Scheduling services define how the BS allocates uplink and downlink transmissions of PDUs. The following scheduling services are supported:

  • UGS
  • ertPS
  • rtPS
  • nrtPS
  • Best Effort

Multiple levels of allocations are supported for ertPS. In the Physical Layer Enabled efficiency mode, connections can request increasing levels of allocation based on their increased load (and decreasing levels for decreased loads). In the Mobility and Ranging Enabled efficiency mode, a connection can also allocate no resources if it has no load. If resources are needed later, the connection uses CQI code to quickly reallocate resources based on the new load.

Bandwidth Request Mechanisms

The SSs use bandwidth request mechanisms to request bandwidth for their uplink transmissions. The following types of bandwidth request mechanisms are supported:

  • CDMA contention-based requests are modeled as aggregated requests.
  • Piggybacked bandwidth requests are modeled as incremental requests. Piggybacked bandwidth requests are disabled by default, but you can enable this feature on any SS by toggling the attribute WiMAX Parameters > SS Parameters > Piggyback BW Request.

Modulation and Coding

Two types of coding schemes are supported

  • Convolutional turbo coding (default scheme)
  • Convolutional coding

For OFDMA, the following modulation and coding schemes are supported: QPSK 1/2, QPSK 3/4, 16-QAM 1/2,16-QAM 3/4, 64-QAM 1/2, 64 QAM 2/3, 64 QAM 3/4, and 64 QAM 5/6. 64 QAM 5/6 is supported for convolutional turbo coding only.

Repetition coding is supported only for QPSK modulation with convolutional coding 1/2 and 3/4.

IPv4 and IPv6 Support

The WiMAX model is completely compatible with IPv4 and IPv6.

Broadcast and Multicast Traffic

The model supports broadcast and multicast traffic when physical layer modeling is enabled. WiMAX treats broadcast and multicast traffic from upper layers as broadcast traffic.

Reserved Subframe Capacity

Base station nodes include attributes that let you model a pre-allocated utilization of the uplink or downlink subframe. The subframe capacity is reduced by the percentage value specified. Use the following attributes to model reserved frame capacity:

  • WiMAX Parameters > BS Parameters > Reserved DL Subframe Capacity
  • WiMAX Parameters > BS Parameters > Reserved UL Subframe Capacity

Overbooking and Underbooking for Admission Control

You can specify that admission control admits more or fewer connections than the subframe capacity (for UGS and polling services). Two attributes on BS nodes, UL Loading Factor and DL Loading Factor, let you overbook or underbook the number of connections.

Overbooking and underbooking are useful for bursty traffic that is not being sent at the same time. Overbooking allows you to admit more connections than the frame capacity and operates under the assumption that not all connections will be transmitting at the same time. Underbooking artificially reduces capacity, which leaves room for best-effort traffic that might otherwise might get preempted.

Automatic Retransmission Request (ARQ) Support

The model includes the following support for ARQ, a retransmission mechanism that resides in the WiMAX MAC layer:

  • Per-connection support for rtPS, nrtPS and BE connections
  • Ability to enable ARQ for individual service flows at subscriber station nodes
  • Ability to enable in-order delivery of SDUs
  • Ability to enable CRC-32 overhead for individual service flows. CRC-32 is enforced whenever ARQ is modeled in a service flow.
  • Selective acknowledgements. Two types of selective acknowledgements are supported: bit sequence maps and block sequence maps.

Hybrid Automatic Retransmission Request Support

The following features are implemented for hybrid ARQ (HARQ):

  • Chase combining at the physical layer
  • Support for polling services, best effort and ertPS type of traffic scheduling
  • Higher priority retransmissions in the MAC scheduler
  • Intelligent scheduler for HARQ channel management and granting retransmissions and acknowledgements
  • Implicit acknowledgement for the uplink channel and explicit acknowledgement for the downlink channel

Note that unlike ARQ, HARQ can be activated on ertPS service flows as well—but not on UGS service flows. HARQ is allowed on ertPS service flows because ertPS has efficient allocation if there is a silence period in the traffic flow using ertPS. Because of this, HARQ retransmissions can use the extra bandwidth provided by the silence periods and still remain in sync. When configuring HARQ, be sure to configure the correct MRR for ertPS service flows by noting the traffic load and accounting for HARQ retransmissions that may occur.

The decision to use ARQ or HARQ depends on the type of environment used in the simulation. While both ARQ and HARQ achieve the same goal of improving packet delivery ratio, there could be some trade-offs. The first trade-off is that both ARQ and HARQ have unpredictable retransmissions, which are handled by the MAC scheduler. If there are too many retransmissions, the retransmissions would use most of the bandwidth and cause very high delays. HARQ is better at obtaining lower delays because by design, acknowledgement comes in a few frames at most. ARQ on the other hand, must wait a long time before it realizes that packets are lost.

ARQ and HARQ can both be used together. The recommended configuration is to use both with their default values. With this configuration, HARQ is the primary source of retransmissions due to its fast response times. If HARQ fails after reaching the maximum retransmission attempts, ARQ kicks in and can do another round of MAC retransmissions, which will ultimately involve HARQ again. Essentially, if ARQ is slow and HARQ is fast, ARQ acts as a higher-layer protocol for HARQ.

Unlike ARQ, HARQ can be activated on ertPS service flows as well—but not on UGS service flows. HARQ is allowed on ertPS service flows because ertPS has efficient allocation if there is a silence period in the traffic flow using ertPS. Because of this, HARQ retransmissions can use the extra bandwidth provided by the silence periods and still remain in sync. When configuring HARQ, be sure to configure the correct MRR for ertPS service flows by noting the traffic load and accounting for HARQ retransmissions that may occur.

Power Savings Mode

This feature allows mobile subscribers to consume less power during periods of low traffic activity. Power saving is achieved by turning off parts of the MS in a controlled manner when it is not actively transmitting or receiving data. This feature is configurable in the Power Savings Parameters attribute on subscriber stations (SS), also known as Mobile Stations (MS). The MS is modeled to be in one of three states:

  • Normal operations—In this state, the behavior of the MS is invariant.
  • Active/Listening—In this state, the MS will not transmit any messages. It can only receive "broadcast/multicast" messages from the Base Station (BS). The MS can also perform scanning by taking measurements from neighboring BSs.
  • Sleep—This is a total shutdown state, and the MS is unavailable for any transmission/reception or monitoring activity.

Note: Power saving performance might degrade with high frame utilization.

The model supports the following Power Saving Classes:

  • Type-1 Power Saving Class—This class is available for service flows using nrtPS/BE scheduling. In this class, the sleep window is exponentially increased from a minimum value to a maximum value when no traffic exists. In case of any traffic activity, the sleep window is set back to its minimum value.
  • Type-2 Power Saving Class—This class is available for service flows using rtPS/ertPS/UGS scheduling. In this class, the sleep window is fixed length and the sleeping pattern depends on the Traffic Rate that is defined on the MAC Service Class Definitions table on the WiMAX Configuration Object.

Relay Nodes (Based on 802.16j)

802.16j supports the relay of WiMAX packets (control and data) within the same WiMAX cell. A new node, called a relay station (RS) facilitates this function. A transparent RS operating in centralized mode for uplink only is supported. Note that there is no increase of coverage of WiMAX cells with a relay of this type; however, using relays may decrease the transmission power consumption of SSs associated with the relays. Relays can also increase the cell capacity by allowing MSs associated with the relays to use better modulation and coding schemes (MCS). For example, without an RS, a distant SS must either use a very conservative MCS or increased transmission power to reach the BS on the uplink (UL). Having an RS, however, can alleviate both issues. The BS can use the information is has to set a lower initial power through initial ranging of the SS, if it is closer to an RS. The BS can also assign an aggressive MCS if it knows that the SS is close to an RS. The trade-off is that the BS has to make two allocations instead of one: one on the access part of the uplink frame and one on the relay part of the uplink frame.

UL Burst Transmission Delay

The UL Burst Transmission Delay feature allows you to choose whether the uplink grant sent in frame n is for transmission of uplink data in frame n or in frame n + 1.Enable this feature by setting the value of the WiMAX Parameters > BS Parameters > UL Burst Transmission Delay attribute to “One Frame” on the base station.

本站文档由 北京智网畅联科技有限公司 整理、翻译并提供 · Documentation compiled and provided by Beijing ZNetLink Technology Co., Ltd.

交流群 · QQ Group:OPNET/智网妙算网络仿真交流群(群号 486383494)

原文档 © 2017 Riverbed Technology · 本站为公益翻译,由 AI 辅助完成,请仔细甄别

Original documentation © 2017 Riverbed Technology. Community translation, AI-assisted. Contact: jiazequn@znetlink.com