Wi-Fi Modelling

• Modelling and Simulation of Physical and MAC Layer for WLAN (2.4 GHz & 5 GHz) for IEEE 802.11n & 802.11ac.
• MAC layer: Rate adaptation algorithms
• Coexisting WLAN networks
• Downlink Performance of OFDMA 802.11ax WLAN Networks in dense deployments.

Standards considered: 802.11n & 802.11ac at 2.4 GHz and 5 GHz

Performance metrics

–Throughput
–Latency
–Packet Error Rate
–Range

Models for various scenarios

–Scenario 1: Client at different distances from an Access Point
–Scenario 2: Client in motion WRT a single Access Point
–Scenario 3: Client in motion while covering multiple Access Points

• Rate adaptation is used to deal with variations in channel and interference
  • Transmitter adapts its modulation and coding scheme (MCS) to get different rates
• A crucial algorithm that affects performance of Wi-Fi networks
• Factors that affect throughput:
  • Dynamic nature of channel
  • Interference from other WiFi transmitters
  • Heterogeneous receiver capability
• Goal: Maximize throughput while containing packet error rate

Comparison of DRCA, Ideal RAA, and Minstrel:

Matlab’s Dynamic Rate Control Algorithm (DRCA):

Ideal Rate Control Algorithm:

Benchmarking: Ideal vs. DRCA (802.11ac):

Minstrel-HT Rate Control Algorithm:

• Evaluating the Matlab implementation of Minstrel-HT with a reference known as MinstrelHTWifiManager from NS3.

• Response of algorithm (MCS selection and MRR update) to packet failures/success given the same initialization and randomisations were matched.

• WLAN co-existence analysis inside a plane
  • Analysing throughput between interfering APs

• Minstrel-HT integration with multi-node system-level simulations

• Simulated Rate adaptation in an increased number of WLAN networks

• Gain of the MU-OFDMA feature of the 802.11ax standard provide
• 11ax vs 802.11n simulations: Implemented in Matlab
  • 11n vs. 802.11ax
  • Throughput-based scheduling for extreme fairness

• Scheduled access as opposed to contention-based access.
• Different RUs for different STAs depending on channel conditions.

• 11ax: OFDMA for MU transmissions
  • Compare per-node throughput and latency

• Round-robin vs. throughput-based OFDMA scheduler

• Default round-robin scheduler in Matlab:
  • RUs allocated based on the number of previous transmissions.
  • STAs with the least number of transmissions at the start of each frame are allocated RUs.
• Throughput-based scheduler:
  • RUs allocated based on the running value of throughput.
  • STAs with the least throughput at the start of each frame are allocated RUs.
• Scheduled access vs. contention-based access:
  • Allocating RUs based on SINR per RU for each STA.

Other Team members: Amol Bhindwale, Rushabha Balaji, Sourav Mukherjee, Vinay U Pai