Getting the best out of your link budget – channel aware rate selection
One of the most important aspects in ensuring a controlled and predictable QoS level over the ever changing wireless medium is an optimized over-the-air rate selection scheme. Such a scheme should on one hand maintain a low packet error rate to reduce retransmissions and increase air time efficiency, and on the other hand choose the highest possible optimal rate to maximize throughput. All this needs to be done over a fast changing wireless medium.
Traditionally, APs have selected the OTA rate based on slowly reacting channel-state-ignorant PER (Packet Error Rate) based measurements that necessitate the gathering of long term statistics. A video-grade AP should endeavor to base its decisions on a reliable estimate of the actual instantaneous signal-to-noise ratio at the client device. This enables extracting the maximum capacity out of the wireless link with a controlled and predictable packet error rate, in any channel condition, which in turn guarantees controlled delay and jitter to enable robust HD streaming.
Traditional slowly reacting PER based rate selection methods are prone to long bursts of packet retransmissions in cases of fast channel fade. These retransmissions significantly increase the underlying traffic jitter to which the HD video signal is highly sensitive.
On the other hand, when the channel state rapidly improves, a slow reacting rate selection scheme will fail to make optimal use of it. In fact, PER-based rate selection schemes usually employ large safety margins to avoid the above mentioned retransmission bursts, and hence can never fully exploit the available channel capacity.
Figure 3: Rate Selection
The instantaneous channel capacity can easily be tracked by monitoring the changing SNR conditions. One particularly useful way to monitor SNR in the case of explicit beamforming in 802.11n is to use the SNR metrics that accompany the channel state information feedback from the client device.
Scheduling of Multiple HD Streams and Fairness Mechanisms
The scheduler is the master mind that controls when and how each packet is sent over the air and assigns priorities to the different streams to enable robust operation on what is otherwise an unreliable wireless medium.
The scheduler should combine channel awareness and knowledge of outgoing video queue status for each client to make optimal transmit decisions that ensure optimal usage of the air medium and maximum MAC efficiency. This is done by making frame aggregation and rate selection decisions that ensure the per-stream delay and jitter are minimized and bounded, to ensure in turn a robust transmission of jitter sensitive HD streams.
The channel-aware scheduler is also responsible for scheduling channel sounding frames to provide for fresh beamforming vectors for all clients, and for gathering performance and interference metrics and statistics for accurate channel tracking.
One of the most important tasks of the scheduler, which is particularly important for an HD home network supporting portable devices, is to maintain fairness between all clients with appropriate admission control. This way one client device will not consume too much air time to the detriment of the other devices. This situation can arise when one of the video client devices is very far from the AP, for example, and needs to be served in a relatively low air rate.
Assuming all clients receive an HD stream of similar bandwidth, this remote client will consume most of the air time, and “rob” the other clients from their fair share of medium resources. If this situation is not controlled, all clients will suffer video signal degradation, even though only one remote client can have a bad channel state. In this case, the scheduler should limit the air time of the remote client such that it will not degrade other client’s performance.Each client performance will be gracefully degraded according to its own channel condition without affecting the other client devices.
The time has come for mature video-grade Wi-Fi solutions that can support multiple HD video portable devices. Using channel aware cross-layer optimization of transmit parameters like OTA-rate, beamforming vectors, antenna switching, and scheduling can make the difference and achieve robust and predictable video delivery over otherwise hostile wireless channels.
Figure 4: Channel Aware Silicon from Celeno
About the Author
Nir Shapira, CTO Celeno
. Nir Shapira brings to Celeno over two decades of experience in the fields of communication theory, signal processing and wireless communications. Nir joined Celeno from the day of its inception and was responsible for forming the technological infrastructure and intellectual property of the company. He also represents Celeno in IEEE 802.11 standardization activities. Prior to joining Celeno, Nir held various managerial roles in Conexant, developing state-of-the-art wireless, DSL and voice band communication technologies. He served in an elite R&D unit of the IDF. Nir received a B.Sc. in EE and Physics (summa cum laude) from the Technion, Israel’s Institute of Technology, and an MBA from the Tel-Aviv University.
If you liked this article...
- Head to the RF and Microwave Designline homepage for the latest updates in RF and microwave technology.
- Sign up for the weekly RF and Microwave Designline Newsletter to be delivered to your mailbox with the latest highlights from the site.