Wireless Access to Ultimate Virtual Reality 360-Degree Video At Home

12/05/2018 ∙ by Huanle Zhang, et al. ∙ University of California-Davis Simula Research Lab 0

Virtual reality 360-degree videos will become the first prosperous online VR application. VR 360 videos are data-hungry and latency-sensitive that pose unique challenges to the networking infrastructure. In this paper, we focus on the ultimate VR 360 that satisfies human eye fidelity. The ultimate VR 360 requires downlink 1.5 Gbps for viewing and uplink 6.6 Gbps for live broadcasting, with round-trip time of less than 8.3 ms. On the other hand, wireless access to VR 360 services is preferred over wire-line transmission because of the better user experience and the safety concern (e.g., tripping hazard). We explore in this paper whether the most advanced wireless technologies from both cellular communications and WiFi communications support the ultimate VR 360. Specifically, we consider 5G in cellular communications, IEEE 802.11ac (operating in 5GHz) and IEEE 802.11ad (operating in 60GHz) in WiFi communications. According to their performance specified in their standards and/or empirical measurements, we have the following findings: (1) Only 5G has the potential to support both the the ultimate VR 360 viewing and live broadcasting. However, it is difficult for 5G to support multiple users of the ultimate VR live broadcasting at home; (2) IEEE 802.11ac supports the ultimate VR 360 viewing but fails to support the ultimate VR 360 live broadcasting because it does not meet the data rate requirement of the ultimate VR 360 live broadcasting; (3) IEEE 802.11ad fails to support the ultimate VR 360, because its current implementation incurs very high latency. Our preliminary results indicate that more advanced wireless technologies are needed to fully support multiple ultimate VR 360 users at home.

READ FULL TEXT VIEW PDF
POST COMMENT

Comments

There are no comments yet.

Authors

page 1

page 2

page 3

This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.

I Introduction

As the next-generation Internet platform, Virtual Reality (VR) has aroused great interest in academia and industry. The global VR market was valued at USD 3.13 billion in 2017 and is expected to reach USD 49.7 billion by 2023, at a compound annual growth rate of 58.54% over 2018-2023 [1]

. VR systems will ultimately support human perceptions such as sight, hearing, touch, smell, and taste. Currently, academia and industry are focused on sight and hearing, in the form of VR videos. A VR video system can be further classified into VR 360 (immersive or panoramic videos), free view-point, computer graphics, etc.

VR 360 will become the first prosperous online VR application [2]. It has two forms of applications: viewing and live broadcasting. In VR 360 viewing, a user wears a VR headset that blocks outside view so that the user immerses only on what is being displayed on the headset. The user watches a part of the view, termed as Field-of-View (FoV), while the full-view video is captured horizontally and vertically. In VR 360 live broadcasting, a user uploads live video streams to social media, e.g., Facebook and YouTube, using 360-degree cameras, and meanwhile the video is watched by other online users with VR headsets.

In this paper, we focus on the ultimate VR 360 that satisfies human eye fidelity. An ultimate VR video is encoded with 64 Pixels Per Degree (PPD), 12 bit color-depth and 120 Frames Per Second (FPS). The corresponding video bit rate is 2.3 Tbps! The near-future video compression technique, e.g., H.266, is expected to achieve a compression ratio of 350:1. This reduces the data rate of streaming an ultimate VR 360 video to 6.6 Gbps. A further data rate reduction can be realized in VR 360 viewing by leveraging part-view transmission ( instead of ), resulting in 1.5 Gbps data rate. Therefore, the ultimate VR 360 video requires networking infrastructure that supports 1.5 Gbps downlink for viewing and 6.6 Gbps uplink for live broadcasting, with round-trip time of less than 8.3 ms (refresh interval of 120 FPS).

Home network access support for VR 360 is imperative because VR 360 videos are mostly used at home [2]. There are several challenges of building a home network for the ultimate VR 360, including very asymmetric downlink and uplink demands, multiple VR users at home, and VR 360 user viewing behavior. In addition, wireless access to VR 360 services is preferred over wire-line transmission. The wire-line connections to VR devices not only degrade user experience because the user cannot move freely with a cable connected headset, but also create a tripping hazard since the headset covers the user’s eyes.

Considering the demanding networking requirements, e.g., data rate and latency, of the ultimate VR 360, we explore whether the most advanced wireless technologies from both cellular communications and WiFi communications can support the ultimate VR 360. Specifically, we consider 5G in cellular communications, IEEE 802.11ac (operating in 5GHz) and IEEE802.11ad (operating in 60GHz) in WiFi communications. For each wireless technology, we survey its development/deployment status and technology features, as well as the performance determined by its standard specification and/or empirically measured.

Fig. 1: Illustration of VR 360 video size in the early stage, the entry-level experience stage, the advanced experience stage and the ultimate experience stage [2]. Each block represents 30 frames per second. As of late 2018, only the early stage is fully supported

We have the following findings: (1) Only 5G has the potential to support both the the ultimate VR 360 viewing and live broadcasting. Although 5G has not been standardized yet, it is widely believed that 5G will provide 10 Gbps data rate and 1ms RTT. However, it is difficult for 5G to support multiple users of the ultimate VR live broadcasting at home; (2) IEEE 802.11ac supports the ultimate VR 360 viewing but fails to support the ultimate VR 360 live broadcasting because its data rate is below than the data rate requirement of the ultimate VR 360 live broadcasting. The IEEE 802.11ac specification allows the maximal data rate of 6.9 Gpbs at an AC Access Point (AP) and 3.5 Gbps at an AC client. Our measurement of two AC devices gives 2.3 ms RTT. (3) IEEE 802.11ad fails to support the ultimate VR 360, because its current implementation incurs very high latency. Although IEEE 802.11ad can achieve 6.8 Gpbs at an AD AP and 4.6 Gbps at an AD client, our measurement of two AD devices gives 62.7 ms RTT, which is too long for the ultimate VR 360. Our preliminary results of 5G, IEEE 802.11ac and IEEE 802.11ad indicate that the ultimate VR 360 calls for more advanced wireless technologies that substantially boost data rate and RTT performance. The candidates are 6G (extension to 5G), IEEE 802.11ax (extension to IEEE 802.11ac) and IEEE 802.11ay (extenstion to IEEE 802.11ad).

This paper is organized in this way. First, we elaborate on the networking requirement of the VR 360 (Section II), including different stages of VR 360 products. Then, we explain the challenges to support the ultimate VR 360 at home (Section III). Afterwards, we explore 5G (Section IV), IEEE 802.11ac (Section V) and IEEE 802.11ad (Section VI) one-by-one to determine to what extent each wireless technology supports the ultimate VR 360. We conclude this paper in Section VII.

Ii VR 360 Video Requirement

The industry proposes different stages for VR 360 products, including the early stage, the entry-level experience stage, the advanced experience stage, and the ultimate experience stage [2]. These stages are different with regards to PPD, color depths and frame rates. (1) PPD is related to the display resolution. Given the viewing distance from a display, PPD and the widely used Pixels Per Inch (PPI) are transferable. PPD is independent of the viewing distance, and thus is favored in evaluating VR systems. For a user’s FoV of horizontal degrees by vertical degrees, the length of the horizontal pixels and the vertical pixels is PPD pixels and PPD pixels respectively, with the display resolution of (PPD) (PPD) pixels. It is commonly believed that human fovea is capable of detecting 60 PPD. Therefore, in the proposed ultimate VR 360 [2], 64 PPD is used to safeguard possible perception of discontinuous pixels. (2) Color depth is the number of bits to represent each red, green and blue component. For a system with color depth of , the number of colors that the system can render is considering the combination of the three color components. For example, a video with 8-bit color depth can represent 16.8 million colors, whereas a video with 12-bit color depth can represent 4096 times more. HDMI 1.3 specification defines Deep Color, e.g., 12-bit color depth, in order to “eliminate any potential color banding artifacts that could be seen when there are not enough colors to properly display certain images” [3]. (3) Frame rate is the frequency at which consecutive images (video frames) appear on a display. A high frame rate is required to avoid motion blur. The perception of motion blur depends on many factors such as contrast, brightness, spatial factors, image content. It also varies among different human beings. 120 FPS is gaining popularity as witnessed by the support of AOMedia Video 1 (AV1, successor to VP9) and High Efficiency Video Coding (HEVC, successor to AVC).

The four stages of VR 360 are proposed based on the forecast product evolution. The comparison of the data rate among different stages of VR 360 is illustrated in Figure 1. The early stage VR 360 supports 11 PPD, equivalent to pixels for a full-view frame of each eye; each pixel is encoded by 24 bit (color depth of 8 bit) and the frame rate is 30 FPS. Correspondingly, the video bit rate of the early stage VR 360 is 11.3 Gbps (two eyes). As of late 2018, only the early stage VR 360 is fully supported. For example, HTC Vive Pro that is released in April 2018 renders single-eye view for FoV, which corresponds to 14 PPD. From Figure 1, however, we can clearly see that the early stage VR is toy-ish compared to other stages. In this paper, we focus on the ultimate VR 360 which satisfies human eye fidelity. The ultimate VR 360 adopts 64 PPD, 12 bit color depth, and 120 FPS. The video bit rate of the ultimate VR 360 is 2.3 Tbps!

Fig. 2: Home applications of VR 360, including uplink-hungry VR 360 live broadcasting and downlink-hungry VR 360 viewing. The VR cameras and the VR headsets are connected to the VR 360 servers via wireless connections such as celluar and WiFi

In addition to the extremely high data rate, VR 360 has stricter requirement for networking latency compared to traditional video streaming. In traditional video streaming, whole video frames are transmitted, allowing buffering to alleviate network latency and variation. In VR 360 viewing, however, only the content within user’s FoV are transmitted to the headset, in order to reduce the data rate. For the ultimate VR 360, view transmission is enough [2], which reduce the data rate by 4.5X compared to the full-view transmission (i.e., ). Therefore, part-view transmission is the de facto practice nowadays and also the forseeable future for the VR 360 viewing. However, compared to the full-view transmission, part-view transmission needs to track user head movement and promptly update the region of video content for transmission. Therefore, the Round-Trip Time (RTT) from the generation of head movement tracking to the rendering of FoV views on headset should be smaller than the interval of frame refresh. This requires network RTT to be smaller than  (120 FPS) for the ultimate VR 360.

In practise, compression techniques are used to greatly reduce the data rate of videos. Compared to traditional video streaming, VR 360 videos have higher compression ratio because the left-eye view and the right-eye view are partly overlapped (binocular overlap). Current encoding scheme HEVC (i.e., H.265) is expected to achieve 215:1 compression ratio for VR 360 [2]

. The next-generation encoding scheme, called Future Video Codec (FVC or H.266) is estimated to be ready by 2021 with the goal of 50% more compression over HEVC 

[4]. Therefore, it is reasonable to predict that in the next few years the compression ratio of 350:1 for VR 360 can be realized [2]. With video compression of H.266, the data rate of viewing the ultimate VR 360 (part-view transmission) can be reduced to 1.5 Gbps, and the data rate of the ultimate VR 360 live broadcasting (full-view transmission) can be reduced to 6.6 Gbps.

In summary, a network infrastructure for a single-user ultimate VR 360 requires to support data rate of 1.5 Gbps downlink for viewing and 6.6 Gbps uplink for live broadcasting, with RTT of less than 8.3 ms.

Iii Challenges of Supporting VR 360 at Home

Fig. 3: Timeline for wireless technologies from cellular communications and WiFi communications. We explore 5G, IEEE 802.11ac and IEEE 802.11ad to determine to what extent they support the ultimate VR 360

Home network access support for VR 360 is imperative because VR 360 entertainment video is generally used at home [2]. Figure 2 depicts home applications of VR 360, including downlink-hungry viewing and uplink-hungry live broadcasting. The VR 360-degree cameras and the VR headsets are connected to VR 360 servers through wireless networks, e.g., via cellular base stations, femtocell access points or WiFi access points. It is challenging to build networking infrastructure to support the ultimate VR 360 at home because of the following challenges.

Iii-a Asymmetric Downlink and Uplink

VR 360 viewing and VR 360 live broadcasting have extremely different traffic patterns on downlink and uplink. In VR 360 viewing, the traffic dramatically skews towards downlink like traditional video watching. Although VR 360 viewing needs to continuously track user’s head movement and upload the tracking information to the server, the generated uplink traffic in a few Kbps, is negligible to the downlink traffic that is in Gbps level. On the other hand, VR 360 live broadcasting dramatically skews towards uplink because it uploads captured video frames from a VR 360 camera to a social media platform such as Facebook and YouTube. Therefore, to carry the traffic of both the ultimate VR 360 viewing and live broadcasting, a network is required to support 1.5 Gbps downlink and 6.6 Gbps uplink. With the increasing popularity of online social media sharing, the landscape of home Internet access infrastructure may undergo a radical change: uplink speed is several times higher than downlink speed, in order to support VR 360 live broadcasting.

Iii-B Multiple Users At Home

Home access networks are required to support multiple users at home, in which each user may transmit/receive different VR 360 steams. Users in a household are usually connected to a single access point, e.g., WiFi or femtocell. In the past decade, the average number of people per household in the United States is about 2.5 [5]. The number is expected to remain stable in the next decades. The home access network, therefore, is preferred to simultaneously support three VR 360 streams. Since the ultimate VR 360 requires 1.5 Gbps downlink for viewing and 6.6 Gbps uplink for live broadcasting, the resultant data rate of a typical household with three people can be 4.5 Gbps downlink if they are all viewing VR 360 videos and astonishing 19.8 Gbps if they are all live broadcasting VR 360. In addition to the extremely high data rate, home access networks have difficulty in meeting the network RTT requirement (i.e., less than 8.3 ms), since higher data rate generally causes higher network RTT.

Iii-C VR 360 User Viewing Behavior

According to the VR user behavior report [6] that was published in 2016, for users who bought VR devices in the past year, they spent an average time of 10 minutes in matching VR 360 video everyday. It is expected that the viewing time will increase to 53 minutes in 2020 [2]. Like watching traditional videos, users mostly watch VR 360 video during the leisure time (19:00-23:00). The usage time ratio between the leisure time period and other time period is 8:2, and the ratio is expected to remain unchanged by 2025 [2]. The concentrated usage means that the home network traffic will increase tremendously during the leisure time compared to other time periods because of VR 360. It remains a problem that how a network carrier makes profit from building extremely high-speed home network infrastructure while the usage time is limited. Differentiated pricing for different time periods of networking is still an effective strategy.

Iii-D Wireless Network Access

Untethered high-quality VR 360 experience is highly desirable but challenging. The wire-line connections to VR devices not only degrade user experience because a user cannot move freely with a cable connected headset, but also create a tripping hazard since the headset covers the user’s eyes. On the other hand, wireless technologies have limited throughput and high latency compared to wired-line technologies. For example, we calculate the average RTT of LTE and WiFi profiles from a public dataset [7] that was crowd-sourced in 16 countries. The results show that the average RTT is 246.1 ms and 86.5 ms for LTE and WiFi respectively, which is much larger than the target of 8.3 ms in the ultimate VR 360. With regards to the data rate, LTE networks support theoretical maximum of 326 Mbps and IEEE 802.11n provides theoretical maximum of 600 Mbps, which are also far from the data rate requirement of the ultimate VR 360.

Considering the demanding networking requirements, e.g., data rate and latency, of the ultimate VR 360, we explore whether the most advanced wireless technologies from both cellular communications and WiFi communications can support the ultimate VR 360. Figure 3 shows the timeline for cellular technologies and WiFi technologies. For cellular communications, we consider 5G that is about to be standardized and deployed. For WiFi technologies, we consider IEEE 802.11ac that operates in 5GHz frequency bands and IEEE 802.11ad that operates in 60 GHz frequency bands. Although IEEE 802.11ax and IEEE 802.11ay extends IEEE 802.11ac and IEEE 802.11ad respectively, we do not consider them because they are still under development. For each wireless technology, we survey its development/deployment status and technology features, as well as the performance determined by its standard specification and/or empirically measured.

Iv Wireless Access: 5G

The fifth generation cellular communication, i.e., 5G, is expected to be standardized by the early 2020s. Its design target is to achieve 10-100x peak data rate, 1000x network capacity, 10x energy efficiency, and 10-30x lower latency compared to its predecessor of 4G. Because of the exciting performance that 5G promises, the industry has already been planning to build 5G ecosystems. For example, AT&T plans to deploy 5G networks by the end of 2018, and to reach 19 cities in early 2019; Qualcomm unveiled its 5G millimeter-wave module for smartphones in July 2018, and the first 5G smartphone is expected to come out in 2019 (e.g., LG announced that it will release its 5G smartphones in the first half of 2019).

Iv-a 5G Technology Features

The data rate improvement of 5G over 4G mainly thanks to the following three technology categories: extreme densification and offloading, millimeter wave, and massive Multiple-Input and Multiple-Output (MIMO) [8].

Extreme Densification and Offloading. Making the cells smaller is straightforward but extremely effective to increase the network capacity. It has numerous benefits, including the reuse of spectrum in a given area, and the reduced number of users competing for resources at each cellular Base Station (BS). In principle, cells can shrink almost indefinitely without decreasing Signal-to-Interference Ratio (SIR) of users until every BS serves a single user, as long as the power-law pathloss models hold [9]. This allows each BS to devote its resources to an ever-smaller number of users. A Small Cell is basically a miniature base station and it includes outdoor picocells and microcells, and indoor femtocells.

Millimeter Wave. Millimeter wave (mmWave) refers to frequency bands from 30-300 GHz, where wavelengths are 1-10 mm. It is widely believed that mass market semiconductor technology extends up to 100 GHz in the next years, and thus 5G mmWave focuses on 30-100 GHz. mmWave was deemed unsuitable for wireless communications because of hostile propagation qualities, including high pathloss, oxygen absorption, non-penetration of objects, and also because of high hardware costs. However, mmWave technology is maturing to combat the propagation issues and the hardware costs are falling. In 5G mmWave communication, dual connectivity will be an important feature to prevent loss of coverage, because lower frequency provides better coverage due to lower diffraction loss and improved indoor reach. Therefore, mmWave would be employed for data transmission from small cells while the control plane would operate at microwaves from macro cells. The dual connectivity ensures stable and reliable connections, while providing fast data transmission.

Massive MIMO

. MIMO was introduced in LTE, but at that time it only supports two to four antennas per mobile device and up to eight antennas per BS. A massive MIMO system is typically defined as a system that uses a large number, i.e. 100 or more, of individually controllable antennas. It exploits the high spatial Degrees of Freedom (DoF) provided by the large number of antennas to realize spatial multiplexing, in which data are transmitted to several users on the same time-frequency resource. In addition, a massive MIMO system focuses the signal towards the intended receivers and thus minimizes intra-cell and inter-cell interference. The focusing of signals in a particular direction is made possible by transmitting the same signal from multiple antennas, but with a different phase shift applied to each of the antennas, such that the signals overlap coherently at the intended receiver. Massive MIMO has the promise to provide a substantially increased spectral efficiency per cell.

Iv-B 5G System Performance

(a) Data Rate
(b) RTT
Fig. 4: 5G can support 6 users of ultimate VR 360 viewing and 1 user of ultimate VR 360 live broadcasting

Since 5G has not been standardized yet, its maximum achievable performance is not determined. Mobile and wireless communications Enablers for the Twenty-twenty Information Society (METIS) identifies test cases to represent the problem space that 5G will support [10]. Each test case have different requirements such as data rate, latency, reliability, energy efficiency and availability. For example, METIS defines virtual reality office test case, which maps to 10 Gpbs data rate requirement. 5G is also expected to support the “tactile Internet” [11], which requires RTT of about 1ms. Therefore, it is commonly believed that 5G will at least support 10 Gpbs data rate and 1 ms RTT [8].

Figure 4 shows the estimated performance of 5G in terms of data rate and RTT, and compares its performance with the ultimate VR 360 viewing and live broadcasting. As we can see, 5G can support 6 users of ultimate VR 360 viewing and 1 user of ultimate VR 360 broadcasting. With the network densification by deploying small cells, the total supported users in a given region can still be substantial. However, it is difficult for 5G to support multiple users of the ultimate VR live broadcasting at home.

V Wireless Access: IEEE 802.11ac

IEEE 802.11ac was released in 2013 (Figure 3). It is an evolution from IEEE 802.11n. The first AC router (Netgear R6300) was released in 2012 and the first AC-complaint smartphone (Samsung Mega) came out in 2013. Afterwards, more and more smartphones are equipped with AC WiFi modules. Nowadays, AC functionality is the standard specification of new smartphones. Despite having been on markets for several years, existing AC modules are far from the full capacity of IEEE 802.11ac standard. For example, the specification allows 8 spatial streams, while current AC devices can only support up to 4 spatial streams.

V-a IEEE 802.11ac Technology Features

IEEE 802.11ac improves over IEEE 802.11n in many aspects, including wider channels, 256-QAM support, simplified beamforming, more spatial streams and multi-user MIMO (MU-MIMO) [12].

Wider Channels. IEEE 802.11ac only operates in 5 GHz frequency bands. It introduces 80 MHz and 160 MHz channels, in addition to 20 MHz and 40 MHz that are supported by IEEE 802.11n. Considering that contiguous 160 MHz spectrum might not be available in some areas, IEEE 802.11ac supports two forms of 160 MHz channels: a single contiguous 160 MHz channel or an “80+80 MHz” channel that combines two 80 MHz channels and has the same capacity.

256-QAM Support. IEEE 802.11ac supports BPSK, QPSK, 16-QAM, 64-QAM and 256-QAM, with the latter transmits more bits per symbol while less robust to interference. The new introduction of 256-QAM enables IEEE 802.11ac to support 8 bits per symbol whereas IEEE 802.11n only supports 6 bits per symbol with 64-QAM. The resultant modulation gain is 33.3%.

Simplified Beamforming. Beamforming is a process by which the sender directs its energy towards a receiver to increase the signal-to-noise ratio, and hence the data rate of the transmission. IEEE 802.11n implements beamforming between two devices by negotiating mutually agreeable beamforming functions from the menu of options. Very few vendors implement the same options, and thus there is almost no cross-vendor beamforming compatibility in IEEE 802.11n [12]. Instead of the menu of options methodology, IEEE 802.11ac radically simplifies the beamforming specifications to one preferred technical method.

More Spatial Streams. IEEE 802.11ac increases the maximum number of spatial streams from 4 to 8 at the AP, while the client side remains up to 4 spatial streams compared to IEEE 802.11n. The extra spatial streams of an AP can be used to transmit to multiple clients at the same time. The new introduction of multi-user MIMO (MU-MIMO) represents the greatest potential of IEEE 802.11ac. To keep the transmission separate among multiple users, the AP uses beamforming to focus each of the transmissions towards its respective client. As long as the clients are located in different enough directions, MU-MIMO enables an AP to avoid interference between simultaneous transmissions to clients. Prior to IEEE 802.11ac, all 802.11 standards are single-user. MU-MIMO has the potential to change the way WiFi networks are built because it enables better spatial reuse. Instead of omni-directional transmissions as in the previous WiFi deployment, each AC AP reduces its collision domain to sectors that are directed to clients under communication.

V-B IEEE 802.11ac System Performance

(a) Data Rate
(b) RTT
Fig. 5: IEEE 802.11ac AP supports 4 users of ultimate VR 360 viewing. However, IEEE 802.11ac cannot support the ultimate VR 360 live broadcasting because the data rate of the AC client is below the requirement

Once the channel bandwidth, Modulation and Coding Scheme (MCS), Number of Spatial Streams (NSS) and Guard Interval (GI) are determined, the data rate of AC device can be found by looking up the data rate table (e.g., data rate table in [13]). A full-fledged AC AP supports an 160 MHz channel, MCS of 9, NSS of 8, and short GI, and the corresponding data rate is 6.9 Gbps. The capacity of a full-fledged AC client only differs from an AC AP about the number of spatial streams. An AC client supports 4 spatial streams (NSS=4), with data rate of 3.5 Gpbs. To roughly quantify AC network RTT, we use two PC Engine APU 2 boards with AC modules (model: WLE650V5-18) that are closely placed, and ping each other, which gives an average RTT of 2.3 ms.

Figure 5 shows the system performance of IEEE 802.11ac in terms of data rate and RTT, and compares its performance with the ultimate VR 360 viewing and live broadcasting. The data rate of IEEE 802.11ac is from the standard specification and the RTT is from our measurement. As we can see, both AC APs and clients supports ultimate VR 360 viewing. Specifically, 4 users of the ultimate VR 360 viewing are supported by an AC AP. However, AC fails to support the ultimate VR live broadcasting because its data rate is below the data rate requirement of the ultimate VR 360 live broadcasting.

Vi Wireless Access: IEEE 802.11ad

IEEE 802.11ad was standardized in 2012 (Figure 3). The first AD router (TP-link AD7200) was released in May 2016 and the first AD-complaint smartphone (ASUS Zenfone 4 Pro) came out in September 2017. The market adoption of IEEE 802.11ad is not as successful as IEEE 802.11ac. As of late 2018, the authors are not aware of any other smartphones that are equipped with 802.11ad modules. However, as the first 802.11 standard on millimeter wave communication, IEEE 802.11ad was specifically designed to provide Gbps networking.

Vi-a IEEE 802.11ad Technology Features

IEEE 802.11ad differs from legacy WiFi in many aspects including unique channel propagation behavior, novel beam training, and hybrid MAC channel access [14].

Unique Channel Propagation Behavior. IEEE 802.11ad operates in the 60GHz frequency band, with up to 7GHz unlicensed bandwidth. A typical coverage range of IEEE 802.11ad is within 10 m due to the high signal attenuation. Because of the short-range communication, oxygen absorption plays a minor role, even though it peaks at 60 GHz. IEEE 802.11ad communication is characterized by a quasi-optical propagation behavior where the received signal is dominated by the Line-Of-Sight (LOS) path and reflections from strong reflecting materials. Obstructions such as furniture and human body, can easily break the communication link of 60 GHz. Therefore, IEEE 802.11ad is more suitable to in-room environments where LOS path is available or sufficient reflectors are present.

Novel Beam Training. With quasi-optical propagation behavior, low reflectivity, and high attenuation, beamforming results in a highly directional signal focus. Based on this behavior, the standard introduces the concept of “virtual” antenna sectors that discretizes the antenna azimuth. The beamforming phase in IEEE 802.11ad is split into two sub-phases. First, during the Sector-Level Sweep (SLS), an initial corse-grained antenna sector configuration is determined. The sector alignment information is used in the subsequent optional Beam Refinement Phase (BRP), which fine-tunes the selected sectors. Beamforming training between two devices happens at two different time: before their association when the direction between the two devices is unknown and during the data transmission interval. In addition, the beamforming protocol supports a training procedure for low antenna gain devices and can upload training parameters to a central network coordinator for channel access scheduling.

Hybrid MAC Channel Access. IEEE 802.11ad is intended to support various applications such as wireless PC display that requires real-time uncompressed video streaming, and bulk-file downloading that requires very high data rate. In contrast to legacy WiFi, IEEE 802.11ad adopts a hybrid MAC approach to address different application requirements. Specifically, IEEE 802.11ad incorporates three MAC scheduling: contention-based access, scheduled channel time allocation, and dynamic channel time allocation. (1) Contention-based access follows IEEE 802.11 Enhanced Distributed Channel Access (EDCA), which includes traffic categories to support quality of service, frame aggregation, and block acknowledgements. However, it suffers from the deafness problem because of the used directional antennas. (2) In scheduled channel time allocation access, every Beacon Interval (BI) is dedicated exclusively to a pair of communicating nodes, which provides reliability and the best quality of service. However, it cannot efficiently responds to bursty traffic. (3) Dynamic channel time allocation access is an extension of the IEEE 802.11 Point Coordination Function (PCF). It provides high flexibility in resource allocation and adapts to directional communication. However, the scheduling procedure and the implementation are complex.

Vi-B IEEE 802.11ad System Performance

(a) Data Rate
(b) RTT
Fig. 6: The implementation of AD requires much further improvement to reduce the network RTT in order to support the ultimate VR 360

Similar to IEEE 802.11ac, the standard of IEEE 802.11ad specifies data rates that are supported (refer to the data rate table in [15]). For example, and AD AP achieves its maximum data rate using Orthogonal Frequency-Division Multiplexing (OFDM), 64-QAM, 13/16 code rate, 6 coded bits per single carrier (), 2016 coded bits per symbol (), 1638 data bits per symbol (). The corresponding data rate of an AD AP is 6.8 Gbps. An AD client adopts more energy-efficient transmission of Single Carrier (SC) rather than OFDM, and achieves the maximum data rate of 4.6 Gbps when -16QAM, 4 , and 3/4 code rate are used. To roughly quantify AD network RTT, we ping an AD router (Netgear Nightnhawk X10) from a closely located AD laptop (Acer TravelMate P), which gives an average RTT of 62.7 ms.

Figure 6 shows the maximum system performance of IEEE 802.11ad in terms of data rate and RTT, and compares its performance with the ultimate VR 360 viewing and live broadcasting. The data rate of IEEE 802.11ad is from the standard specification and the RTT is from our measurement. Although AD provides enough data rate for the ultimate VR 360 viewing, the current implementation of AD incurs very high network RTT, which fails to support 8.3 ms RTT as required in the ultimate VR 360.

Vii Concluding Remarks

This paper looks into an emerging application, i.e., VR 360 video. The different stages of VR 360 videos are clarified. Especially, we focus on the ultimate VR 360 which satisfies human eye fidelity. We explore the current most advanced wireless technologies, i.e, 5G for cellular communications, and IEEE 802.11ac and IEEE 802.11ad for WiFi communications, to determine to what extent these technologies support the ultimate VR 360. Our preliminary results show that only 5G can support both the ultimate VR 360 viewing and VR 360 live broadcasting. However, 5G fails to support multiple users of ultimate VR 360 live broadcasting in a household. IEEE 802.11ac supports ultimate VR 360 viewing, but cannot support VR 360 live broadcasting. IEEE 802.11ad cannot support the ultimate VR 360 because its current implementation incurs very high network RTT. None of the current wireless technologies can fully support multiple ultimate VR 360 viewing and VR 360 live broadcasting at home. To fully support ultimate VR live broadcasting in a household with three users, an astonishing 19.8 Gbps is required, with network RTT of less than 8.3 ms. It indicates that we need more advanced wireless technologies, such as 6G (extension to 5G), IEEE 802.11ax (extension to IEEE802.11ac) and IEEE 802.11ay (extenstion to IEEE 802.11ad)

References

  • [1] Mordor Intelligence. Virtual Reality (VR) Market Size, Share, VR Technology, Applications, and Region - Trends, and Forecast (2018 - 2023), https://www.mordorintelligence.com/industry-reports/virtual-reality-market. Accessed on Oct. 27, 2018
  • [2] Huawei Technologies. Whitepaper on the VR-Oriented Bearer Network Requirement. Technical report, 2016.
  • [3] Tom Kopin. Effects of Color Depth. Technical report, Kramer White Paper, 2011.
  • [4] Advanced Television. Stand by for H.266 compression. https://advanced-television.com/2018/02/01/stand-by-for-h-266-compression. Accessed on Oct. 25, 2018.
  • [5] statista. Average Number of People Per Household in the United States from 1960 to 2017. https://www.statista.com/statistics/183648/average-size-of-households-in-the-us/. Accessed on Oct. 27, 2018.
  • [6] mojing. China VR User Behavior Survey (in Chinese). Technical report, mojing, Chinese Institute of Advertisement, Trend Consult, 2016.
  • [7] Shuo Deng, Ravi Netravali, Anirudh Sivaraman, and Hari Balakrish-nan. WiFi, LTE or Both? Measuring Multi-Homed Wireless Internet Performance. In ACM IMC, pages 181-194, 2014.
  • [8] Jeffrey G. Andrews, Stefano Buzzi, Wan Choi, Stephen V. Hanly, Angel Lozano, Anthony C. K. Soong, and Jianzhong Charlie Zhang. What Will 5G Be? IEEE Journal on Selected Areas in Communications, 32(6):1065-1082, 2014.
  • [9] Harpreet S. Dhillon, Radha Krishna Ganti, Francois Baccelli, and Jeffrey G. Andrews. Modeling and Analysis of K-Tier Downlink Heterogeneous Cellular Networks. IEEE Journal on Selected Areas in Communications, 30(3):550-560, 2012.
  • [10] ICT-317669 METIS project. Scenarios, Requirements and KPIs for 5G Mobile and Wireless System. Technical report, METIS, 2013.
  • [11] Gerhard P. Fettweis. The Tactile Internet: Applications and Challenges. IEEE Vehicular Technology Magazine, 9(1):64-70, 2014.
  • [12] Matthew S. Gast. 802.11ac: A Survival Guide – Wi-Fi at Gigabit and Beyond. O’Reilly Media, 2013.
  • [13] IEEE Task Group AC. 802.11ac Part 11 Amendment 4: Enhancements for Very High Throughput for Operation in Bands below 6 GHz. Standard, IEEE, 2013.
  • [14] Thomas Nitsche, Carlos Cordeiro, Adriana B. Flores, Edward W.Knightly, Eldad Perahia, and Joerg C. Widmer. IEEE 802.11ad: Directional 60 GHz Communication for Multi-Gigabit-per-Second Wi-Fi. IEEE Communications Magazine, 52(12):132-141, 2014.
  • [15] IEEE Task Group AD. 802.11ad Part 11 Amendment 3: Enhancements for Very High Throughput in the 60 GHz Band. Standard, IEEE, 2012.