Scheduling strategies and throughput optimization for the Uplink for IEEE 802.11ax and IEEE 802.11ac based networks

The new IEEE 802.11 standard, IEEE 802.11ax, has the challenging goal of serving more Uplink (UL) traffic and users as compared with his predecessor IEEE 802.11ac, en- abling consistent and reliable streams of data (average throughput) per station. In this paper we explore several new IEEE 802.11ax UL scheduling mechanisms and compare between the maximum throughputs of unidirectional UDP Multi Users (MU) triadic. The evaluation is conducted based on Multiple-Input-Multiple-Output (MIMO) and Orthogonal Frequency Division Multiple Acceess (OFDMA) transmission multiplexing format in IEEE 802.11ax vs. the CSMA/CA MAC in IEEE 802.11ac in the Single User (SU) and MU modes for 1, 4, 8, 16, 32 and 64 stations scenario in reliable and unreliable channels. The comparison is conducted as a function of the Modulation and Coding Schemes (MCS) in use. In IEEE 802.11ax we consider two new flavors of ac- knowledgment operation settings, where the maximum acknowledgment windows are 64 or 256 respectively. In SU scenario IEEE 802.11ax throughputs outperform IEEE 802.11ac by about 64% and 85% in reliable and unreliable channels respectively. In MU-MIMO scenario IEEE 802.11ax throughputs outperform IEEE 802.11ac by up to 263% and 270% in reliable and unreliable channels respectively. Also, as the number of stations increases, the advantage of IEEE 802.11ax in terms of the access delay also increases.


Background
The latest IEEE 802.11 Standard (WiFi) [1], created and maintained by the IEEE LAN/MAN Standards Committee (IEEE 802.11), is currently the most effective solution within the range of Wireless Local Area Networks (WLAN). Since its first release in 1997 the standard provides the basis for Wireless network products using the WiFi brand, and has since been improved upon in many ways. One of the main goals of these improvements is to increase the throughput achieved by users and to improve the standard's Quality-of-Service (QoS) capabilities. To fulfill the promise of increasing IEEE 802.11 performance and QoS capabilities, a new amendment, IEEE 802.11ax ( also known as High Efficiency (HE) ) was recently introduced [2]. IEEE 802.11ax is considered to be the sixth generation of a WLAN in the IEEE 802.11 set of types of WLANs and it is a successor to IEEE 802.11ac [3,4]. The scope of the IEEE 802.11ax amendment is to define modifications for both the 802.11 PHY and MAC layers that enable at least four-fold improvement in the average throughput per station in densely deployed networks [5][6][7][8]. Currently IEEE 802.11ax project is in a very early stage of development and is due to be publicly released in 2019 .

Research question
In order to achieve its goals, one of the main challenges of IEEE 802.11ax is to enable simultaneous transmissions by several stations and to enable Quality-of-Service. In this paper we assume that the AP is communicating in a regular fashion with a fix set of stations. We explore some of the UL IEEE 802.11ax new mechanisms given that the AP knows with which stations it communicates and we compare between the unidirectional UDP throughputs of IEEE 802.11ax and IEEE 802.11ac in Single User (SU) and Multi User (MU) modes for 1,4,8,16,32 and 64 stations scenarios in reliable and unreliable channels. This is one of the aspects to compare between new amendments of the IEEE 802.11 standard [9]. The SU scenario implements sequential transmissions in which a single wireless station sends and receives data at every cycle one at a time, once it or the AP has gained access to the medium.

Previous works
Most of the research papers on IEEE 802.11ax so far deal with these challenges and examine different access methods to enable efficient multi user access to random sets of stations. For example, in [10] the authors deal with the introduction of OFDMA into IEEE 802.11ax to enable multi user access. They introduce an OFDMA based multiple access protocol, denoted Orthogonal MAC for 802.11ax (OMAX), to solve synchronization problems and to reduce the overhead associated with using OFDMA. In [11] the authors suggest an access protocol over the UL of an IEEE 802.11ax WLAN based on Multi User Multiple-Input-Multiple-Output (MU-MIMO) and OFDMA PHY. In [12] the authors suggest a centralized medium access protocol for the UL of IEEE 802.11ax in order to efficiently use the transmission resources. In this protocol stations transmit requests for frequency sub-carriers, denoted Resource Units (RU) to the AP over the UL. The AP allocates RUs to the stations which use them later for data transmissions over the UL. In [13] a new method to use OFDMA over the UL is suggested, where MAC Protocol Data Units (MPDU) from the stations are of different lenghs. In [14][15][16][17] a new version of the Carrier Sense Multiple Access with   Collision Avoidance (CSMA/CA) protocol, denoted Enhanced CSMA/CA (CSMA/ECA) is suggested, which is suitable for IEEE 802.11ax . A deterministic backoff is used after a successful transmission, and the backoff stage is not reset after service. The backoff stage is reset only when a station does not have any more MPDUs to transmit. CSMA/ECA enables a more efficient use of the channel and enhanced fairness. In [18] the authors assume a network with legacy and IEEE 802.11ax stations and examine fairness issues between the two sets of the stations.
The rest of the paper is organized as follows: In Section 2 we describe the new mechanisms of IEEE 802.11ax relevant to this paper. In Section 3 we describe the transmission scenario with which we compare between IEEE 802.11ax and IEEE 802.11ac in the SU and MU modes. We assume the reader is familiar with the basics of the PHY and MAC layers of IEEE 802.11 described in previous papers, e.g. [19]. In Section 4 we analytically compute the IEEE 802.11ax and IEEE 802.11ac throughputs. In Section 5 we present the throughput of the various protocols and compare between them. Section 6 summarizes the paper. In the rest of the paper we denote IEEE 802.11ac and IEEE 802.11ax by 11ac and 11ax respectively.
2 The new features in IEEE 802.11ax IEEE 802.11ax focuses on implementing mechanisms to serve more users simultaneously, enabling consistent and reliable streams of data ( average throughput per user ) in the presence of many other users. Therefore there are several new mechanisms in 11ax compared to 11ac both in the PHY and MAC layers. At the PHY layer, 11ax enables larger OFDM FFT sizes, 4X larger, therefore every OFDM symbol is extended from 3.2µs in 11ac to 12.8µs in 11ax. By narrower subcarrier spacing (4X closer) the protocol efficiency is increased because the same Guard Interval (GI) is used both in 11ax and 11ac .
Additionally, to increase the average throughput per user in high-density scenarios, 11ax expends the 11ac Modulation Coding Schemes (MCSs) and adds MCS10 (1024 QAM ) and MCS 11 (024 QAM 5/6), applicable for transmission with bandwidth larger than 20 MHz.
In this paper we focus on UL scheduling methods that enable to optimize the IEEE 802.11 two-level aggregation schemes working point, first introduced in IEEE 802.11n [1,4], in which several MPDUs can be aggregated to be transmitted in a single PHY Service Data Unit (PSDU). Such aggregated PSDU is denoted Aggregate MAC Protocol Data Unit In 11ax and 11ac the size of an MPDU is limited to 11454 bytes. In 11ac an A-MPDU is limited to 1,048,575 bytes and this limit is extended to 4,194,304 bytes in 11ax. In both 11ac and 11ax the transmission time of the PPDU (PSDU and its preamble) is limited to 5.484ms (5484µs) due to L-SIG (one of the legacy preamble's fields) duration limit [1]. The A-MPDU frame structure in two-level aggregation is shown in Figure 1.
IEEE 802.11ax also enables the extension of the acknowledgment mechanism by using a 256 maximum acknowledgment window vs. maximum window of 64 in 11ac. In this paper we also assume that all MPDUs transmitted in an A-MPDU frame are from the same Traffic Stream (TS). In this case up to 256 MPDUs are allowed in an A-MPDU frame of 11ax, while in 11ac up to only 64 MPDUs are allowed.
The acknowledgments are transmitted by special control frames, Block Ack (BAck) and Multi Station BAck, to be specified later.
Finally, in 11ac it is not possible to transmit simultaneously over the UL and only SU is supported. In 11ax this is possible using MU and up to 74 stations can transmit simultaneously.

Transmission patterns
One of the main goals of 11ax is to enable larger throughputs in the network when several BAck frames in return. In this mode the advantage of the 11ax over 11ac is due to its more efficient PHY layer and its new MCSs. The UL traffic pattern in this case is shown in When several stations are transmitting over the UL in 11ac, the air access selection is done by using the CSMA/CA MAC only, which involves collisions between stations, as shown in Figure 2 Figure 3 This format is applicable in 11ax only.

UL Transmissions' service scheduling strategies
In 11ax there are several UL scheduling and non-scheduling service strategies for the stations to transmit data to the AP, and we compare between them. Recall that in 11ac the only possible service strategy is to use the CSMA/CA MAC over the UL, as shown in Figure 2(B).
We now specify the UL service scheduling strategies in 11ax for every number S of stations, S = 1, 4, 8, 16, 32, 64. By SU AX we refer to the traffic pattern in Figure 2(A). By x · SU AX (1) we denote a transmission by n stations in 11ax using the transmission pattern in The UL service scheduling strategies are as follows: • S = 4: 4 · SU AX (1), 1 · MU AX (4).

Channel assignment
We assume the 5GHz band, a 160MHz channel and that the AP and the stations have 4 antennas each. In 11ac every station transmits using 4 SSs. This is because 11ac supports UL SU only and a single station can transmit in all 4 SS if needed. In 11ax a station transmits in SU mode, Figure 2(A) and 2(C), by using 4 SSs and in MU mode by using 1 SS. Recall that in both 11ac and 11ax, and in both SU and MU modes in 11ax, the AP transmits over the DL by using the legacy mode. The DL PHY rate is usually set to the minimum between the UL Data rate and the largest possible PHY rate in the set of the basic rates that is smaller or equal to the UL Data rate. The minimal basic PHY rate is 6Mbps and in the case of UL PHY rates smaller than 6Mbps the DL PHY rate is never less than 6Mbps. This can happen in case of 64 stations (see Table 2).
When using the MU mode in 11ax, the 160MHz channel is divided in the UL into S

PPDU formats
In Figure  In Figure 4 (B) we show the legacy preamble, used in both 11ac and 11ax over the DL.
In Figure 4 (C) we show the PPDU format used in 11ax UL SU mode, Figure 2(A).
In Figure  In the 11ax PPDU format there are the HE-LTF fields, the number of which equals the number of SSs in use, 4 in our case. In this paper we assume that each such field is composed of 2X LTF and therefore of duration 7.2µs [2].
Notice also that the PSDU frame in 11ax contains a Packet Extension (PE) field. This field is mainly used in MU mode and we assume that it is 0µs in SU and the longest possible in MU, 16µs.

Parameters' values
In Table 1 we show the PHY rates and the length of the preambles that are used in 11ac and 11ax in SU mode and in the various MCSs. The values are taken from [2].
In Table 2  We also include again the PHY rates of 11ac in SU mode which are also used when S > 1 stations are transmitting over the UL.
We assume the Best Effort Access Category in which AIF S = 43µs, SIF S = 16µs  · SlotT ime which equals 67.5µs for a SlotT ime = 9µs. Assuming an OFDM based PHY layer every OFDM symbol in IEEE 802.11ac is 3.2µs.
We assume also similar multi-path conditions and therefore set the DL and UL Guard Intervals (GI) to 0.8µs. Thus, in IEEE 802.11ac the duration of a symbol in the DL and UL is 4µs. In IEEE 802.11ax the symbol is 12.8µs. In the DL we assume again a GI of 0.8µs and therefore the symbol in this direction is 13.6µs. In the UL MU we assume a GI of 1.6µs and therefore the symbol in this direction is 14.4µs. The UL GI is 1.6µs due to UL arrival time variants. In SU UL the GI is 0.8µs.
Finally, we assume that the MAC Header field is of 28 bytes and the Frame Control Sequence (FCS) field is of 4 bytes. We also consider several channel conditions which are expressed by different values of the Bit Error Rate (BER) which is the probability that a bit arrives corrupted at the destination. We assume a model where these probabilities are bitwise independent [20].

IEEE 802.11ac
The throughput of 11ac when only one station is transmitting in the network, Figure 2(A), is given by Eq. 1 [19]: where: T (BAck) = T Sym DL · (30 · 8) + 22 T Sym DL · R DL T (DAT A) and T (BAck) are the transmission times of the data A-MPDU frames and the BAck frames respectively. T (BAck) is based on the BAck frame format given in Figure 3 R DL and R U L are the DL and UL PHY rates respectively and P DL and P U L are the preambles used in the DL and UL respectively, see Figure 4.
Concerning the throughput of 11ac where several stations transmit over the UL, we use the analysis in [21] and verify this analysis by simulation.

IEEE 802.11ax
The throughput of 11ax for the MU case, i.e. the traffic pattern in Figure 2(D), is given by Eq. 3 [19]: where: protocol [1]. R DL and R U L are the DL and UL PHY rates respectively and P DL and P U L are the preambles used in the DL and UL respectively (see Figure 4).
The terms in Eqs. 1 and 3 are not continuous and so it is difficult to find the optimal X and Y, i.e. the values for X and Y that maximize the throughput. However, in [19] it is shown that if one neglects the rounding in the denominators of Eqs. 1  We therefore use the result in [19] and look for the maximum throughput as follows: We check for every X, 1 ≤ X ≤ 64 (also 1 ≤ X ≤ 256 for 11ax) and for every Y, 1 ≤ Y ≤ Y max , for the received throughput such that Y max is the maximum

Transmissions' models and scenarios
We compare between all applicable configurations and scheduling flavors of the stations' transmissions up to 64 stations. The scheduling flavors are as follows.
Concerning 11ac : • UL using CSMA/CA . DL Ack transmissions are conducted at the basic rate set. i.e. we look for the optimal A-MPDU frame structure.
In the next section we show three sets of results. In Figure 5

Throughput results
Recall that in Figure 5  For 11ac we show analytical results received from the analysis in [21] and we also verify these results by simulation. For 11ax, when there is only one station in the network we use We see in Figure 5(A) that the largest throughput is received in SU AX (1). Notice however that the throughput of SU AX (1) when only one station is transmitting in the system is larger than the throughput of SU AX (1) when S > 1 stations are transmitting. This is due to the lack of the TF frame when one station transmits, and using the BAck frame which is shorter than the Multi Station BAck. SU AX (1) has the largest throughput among all transmission scheduling flavors because of its relatively larger PHY rate -it is larger than 4 times the one in the case of 4 stations, larger than 8 times the one in the case of 8 stations etc.
The throughout of MU AX (8) is the same as that of MU AX (4). From Table 2  The throughput of MU AX (64) is the smallest because of its very small PHY rates which are much less than half those in MU AX (32). Recall also that MCS10 and MCS11 are not applicable in the case of 64 stations. Also, the transmission of the TF frame now requires 7 symbols.
Finally, 11ax outperforms 11ac by 78% and 263% for 4 and 64 stations respectively because 11ax uses a scheduled transmission pattern while 11ac is based on the contention CSMA/CA MAC protocol access with collisions.
Although the throughput metric is important, the access delay metric is also important.
This metric is defined in this paper as the time elapsed between two consecutive transmissions from the same station to the AP.
In Figure 5 In overall MU AX (16) and MU AX (32) seem to be the best transmission scheduling flavors achieving large throughputs with small access delays.
In Figure 6 we show the throughput performance of MU AX (4) and MU AX (64) for every MCS, for BER=0 and 10 −5 , and for the cases using 64 and 256 MPDUs per A-MPDU frame.
In         In this paper we explore multiple scheduling strategies in order to compare between the throughputs of 11ac and 11ax over the Uplink when considering UDP traffic and that several stations are transmitting in the system. There is an optimal working point for every scheduling strategy in terms of the A-MPDU frame structure. In MU AX (4) it is sufficient to transmit around 70 and 256 MPDUs per A-MPDU frame for BER=0 and BER=10 −5 respectively. For MU AX (64) these numbers of MPDUs are smaller, around 10 and 40 respectively, due to the smaller PHY rates.
Finally, using up to 256 MPDUs in an A-MPDU frame outperforms the use of up to 64 MPDUs in cases when the PHY rates are larger and/or the channel is unreliable, i.e. BER=10 −5 .