TACTILE INTERNET IN THE BEYOND 5G ERA

The TI is centered on the H2M interactions with the haptic devices. To facilitate haptic communication, the
transmission of data via tactile internet creates a network that is both extremely reliable and extremely responsive.
The tactile internet’s purpose is to provide a remote and dynamic way for people to experience physical haptic or the
touch based control, while exchanging closed-loop information between the virtual and physical worlds. Wireless
communications can thus be a medium for controlling and directing real and virtual objects using such a platform.
[1]

This revolutionary technology continues to transform healthcare, transportation, education, logistics, smart grid
systems and many more, hence covering a major portion of the economy sector in the society. It thus provides sub-
millisecond connectivity for the healthcare applications like remote surgery. With high availability of TI,
accompanied with its very fast reaction time and reliability, the human interaction with the machines enables a new
dimension by creating an interactive, real-time system, which revolutionizes almost every segment of the society.[1]

TACTILE INTERNET ARCHITECTURE

In contrast to the traditional Internet, which is used to transmit audio and video data, TI intends to transmit touch
and actuation data in addition to audiovisual data. TI systems can include kinesthetic and vibro-tactile parameters
like position, motion, vibration, and surface texture in their feedback, creating a global control loop, while
traditional systems only use audio/visual feedback. Also, the haptic sense in TI applications works both ways. For
example, in a telepresence system, the motion (i.e., velocity or position) is sent to the tele-operator and the
force/torque from the environment is sent back to the Human System Interface (HSI).[2]

A TI system's general architecture, which mainly comprises a master domain, a network domain, and a controlled
domain. The master domain is typically comprised of a human operator (tactile user) and the HSI, which
is responsible of converting human input to tactile data via appropriate tactile encoding techniques.[3] The
controlled domain or environment contains a remotely controlled robot or teleoperator, and the master domain
directly controls the controlled domain using various command signals (e.g., velocity, position).
The network domain of the TI system serves as the channel of communication between tactile users (master domain)
and the remotely controlled environment (controlled domain). This communication domain may be composed of  an
Internet/core network, a Radio Access Network (RAN), and a tactile support engine. The communication channel
must also be ultra-responsive and ultra-reliable. B5G networks should be modified to support emerging TI
applications. B5G core network supports TI systems by managing edge-cloud interactions, application-aware QoS
provisioning, and security.[3]

Figure 1
KEY TECHNICAL REQUIREMENTS

The following are some of the most important technical requirements for the TI:

Ultra-responsive Connectivity. Most TI applications require 1 ms round-trip latency. End-to-end

latency (round-trip delay) in technical systems is the summation of sensor/device transmission
times (or human for haptic communication). Via the communication infrastructure to a control
server, information processing at the server, processing at multiple communication intermediate
nodes (i.e., routers, switches), and retransmission times back to the end-device (or human).[3]

Ultra-reliable Connectivity: The TI also needs ultra-reliable network connectivity, which means
the probability to ensure a required performance under complete structural constraints and
conditions during a specific time interval. One way to improve the reliability of TI applications is
to use simultaneous connections with multiple links and multiple paths for graph connectivity in
order to avoid single point of failure. Improving reliability will also help to reduce latency due to
the lower number of resulting retransmissions.[4]
Distributed Edge Intelligence: In order to enable the interpolation and extrapolation of human
activities and predictive caching for lowering end-to-end latency, appropriate Artificial
Intelligence (AI)/ML techniques need to be investigated for use at the edge-side of the wireless
TI networks.[3]
Transmission and Processing of Tactile Data. Transmission of tactile information over packet-
switched networks requires the development of tactile encoding mechanisms. Furthermore, an
efficient audio/visual sensory feedback mechanism must be investigated in order to deal with the
highly multidimensional nature of human tactile perception.
Security and Privacy: Under strict latency constraints, TI security and privacy are extremely
important. To meet these requirements, low-computational overhead physical layer security,
secure coding, and reliable and low-latency methods to identify authentic receivers need to all be
investigated.[3]

APPLICATION AREAS

As previously stated, the TI is expected to have an impact on various aspects of human society,
with applications ranging from eHealth Care to industrial automation. Among the application
scenarios described in the literature are: Haptic communications, Augmented Reality
(AR)/Virtual Reality (VR), Remote monitoring and surgery, Wireless controlled exoskeletons,
Remote education (telementoring), Remote driving/autonomous vehicles, Traffic control,
Industrial tele-operation such as robotics and manufacturing, Smart grid and Smart city.[3]
The baseline TI standard IEEE 1918.1 lists the key features and performance metrics of the
following use cases: Immersive VR (IVR), teleoperation, automotive, Internet of drones,
interpersonal communications, live haptic-enabled broadcast, and cooperative automotive
driving.

FEASIBILITY ANALYSIS OF EXISTING WIRELESS TECHNOLOGIES

Existing wireless technologies and protocols, which are primarily designed for conventional
HTC traffic, are insufficient to meet the stringent latency and reliability requirements of
emerging TI applications requiring M2M and M2H interactions. Therefore, it is critical to
analyze the feasibility of existing technologies/protocols for TI application scenarios and to
explore adaptations of existing protocols for TI applications.[5]

Some recent studies have examined the feasibility of a few TI applications in different settings,
as briefly discussed below
.
Factory automation is an example of a TI use case that has very strict requirements for latency
and reliability, such as an end-to-end latency of less than 1 millisecond with a failure rate as low
as 10−9.In this context, the authors of [3]examined the viability of developing a 5G radio
interface based on Orthogonal Frequency Division Multiplexing (OFDM) for a time-sensitive
MTC application. The authors used numerical results to stress the importance of a new 5G radio
interface for applications requiring 99.999% reliability and less than 1 ms of end-to-end latency.

The video communications system is another TI application scenario developed to accommodate

NCSs, interactive applications that are constantly evolving, and new forms of cyber-physical
networking. Glass-to-glass delay refers to the time it takes for photons of an event to travel from
a camera lens to a monitor, which is the end-to-end delay in video communications.[3]

Due to time-varying channel fading and interference, providing reliable wireless communications becomes
challenging. To satisfy stringent reliability requirements, signal quality outage performance can be studied by
analyzing suitable performance metrics such as Signal to Noise plus Interference (SINR) outage. . In [3], authors
performed a signal quality outage performance analysis of a cellular network and discovered that conventional
Multiple-Input Multiple-Output (MIMO) systems with 2 × 2 or 4 × 4 antenna configurations are not sufficient  the
stringent reliability requirements. To improve this performance, the same authors proposed using macroscopic
diversity as well as interference management technique is compared with MIMO schemes to improve a cellular
network's SINR outage performance. However, when employing the discussed MIMO diversity techniques, the
diversity-multiplexing tradeoff issue must be properly considered.

In [3], authors examined the feasibility of IEEE 802.11 (Wi-Fi) technology for low-latency TI applications using
both DSSS and OFDM schemes. SDR-based testbeds were used to simulate the effects of concurrent timing offset
and concurrent frequency offset from physically uncoupled devices used for simultaneous Wi-Fi transmissions. The
results showed that the IEEE 802.11 DSSS and IEEE 802.11 OFDM systems behaved differently when concurrent
transmissions limited them. The main IEEE 802.11 parameter limits for ultra-low-latency TI applications were
identified.

ENABLING TECHNOLOGIES FOR WIRELESS TACTILE INTERNET

A. REVIEW OF MAIN CHALLENGES
Wireless networks across TI application domains.
The main connectivity challenges are ultra-low latency of < 1 ms, ultra-high reliability of >
99.99%, high data-rate for some TI applications in the order of Gbps to Tbps, and very high
backhaul bandwidth [3] Wireless Extended Reality (XR), wireless AR/VR, has additional
challenges beyond connectivity, such as balancing the quality-rate-latency trade-off, dealing with
scalability and heterogeneity, improving localization and tracking efficiency, choosing between
in-VR and in-network computation, and investigating novel information theoretic principles to
characterize wireless XR systems [3]

Providing prescriptive QoS for TI services in 5G and beyond wireless networks is a major
research challenge. The main delay components, such as transmission delay, processing delay,
and queueing delay, should have bounds with a low probability of violation [6]. Another
challenge is achieving the Quality of Service (QoS) requirements of TI services without
compromising the spectral efficiency or energy efficiency standards necessary for 5G and future
networks.

In addition, the majority of the currently available wireless systems were developed to
accommodate Human-Type Communications (HTC) traffic, and so they may not be well-suited
to effectively manage massive number of short data packets produced by IoT devices in TI
environments [3].
Furthermore, many existing congestion control mechanisms, higher layer and link layer based
retransmission techniques for improving the communication reliability may not be applicable in
TI applications due to the challenging requirements (mainly due to the strict requirement of 1 ms
round-trip latency).
B. ENABLING TECHNOLOGIES
Table shows the key techniques for supporting TI applications in 5G and beyond networks and
their sub-methods.
table
Protocol Layer Enabling Technologies Sub-methods
Cross-layer resource optimization
Energy-efficient resource allocation
Haptic resource allocation
Dynamic resource allocation Multi-cell resource allocation
techniques Joint uplink-downlink resource allocation
Traffic aware resource allocation
Physical and MAC Massive MIMO
Layer Channel reciprocity
Advanced signal processing Area spectral efficiency optimization
techniques Bustiness-aware Bandwidth Reservation
Diversity for transmission reliability
Coding and modulation
Asymmetric Transmit-Windowing
Transmission and Link adaptation Multicarrier waveforms
schemes HCCA MAC protocol
Control channel design
Non-orthogonal Multiple Access (NOMA)
Pilot/overhead minimization
Latency reduction methods
Reliability enhancement techniques
Multiple access and scheduling
techniques

Latency reduction and reliability

enhancement
Software-Defined Networking
(SDN)
Network Layer and Network virtualization
Cloud-Level Network coding
Collaborative edge-cloud processing
Caching techniques
Machine learning/AI
Network architectural aspects
Key technical requirements.
key application areas.
TI architecture.
the main enabling technologies.
Key technical requirements.

Key technical requirements.

Ultra-responsive Connectivity.
Most TI applications require 1 ms round-trip latency. End-to-end latency (round-trip delay) in
technical systems is the summation of sensor/device transmission times (or human for haptic
communication). via the communication infrastructure to a control server, information
processing at the server, processing at multiple communication intermediate nodes (i.e., routers,
switches), and retransmission times back to the end-device (or human).
Ultra-reliable Connectivity.
The TI also needs ultra-reliable network connectivity, which means the probability to ensure a
required performance under complete structural constraints and conditions during a specific time
interval. One way to improve the reliability of TI applications is to use simultaneous connections
with multiple links and multiple paths for graph connectivity in order to avoid single point of
failure. Improving reliability will also help to reduce latency due to the lower number of
resulting retransmissions.
Distributed Edge Intelligence:
In order to enable the interpolation and extrapolation of human activities and predictive caching
for lowering end-to-end latency, appropriate Artificial Intelligence (AI)/ML techniques need to
be investigated for use at the edge-side of the wireless TI networks.
Transmission and Processing of Tactile Data.
Transmission of tactile information over packet-switched networks requires the development of
tactile encoding mechanisms. Furthermore, an efficient audio/visual sensory feedback
mechanism must be investigated in order to deal with the highly multidimensional nature of
human tactile perception.
Security and Privacy:
Under strict latency constraints, TI security and privacy are extremely important. To meet these
requirements, low-computational overhead physical layer security, secure coding, and reliable
and low-latency methods to identify authentic receivers need to all be investigated.

APPLICATION AREAS
As previously stated, the TI is expected to have an impact on various aspects of human society,
with applications ranging from eHealthCare to industrial automation. Among the application
scenarios described in the literature are: Haptic communications, Augmented Reality
(AR)/Virtual Reality (VR), Remote monitoring and surgery, Wireless controlled exoskeletons,
Remote education (telementoring), Remote driving/autonomous vehicles, Traffic control,
Industrial tele-operation such as robotics and manufacturing, Smart grid and Smart city.
The baseline TI standard IEEE 1918.1 lists the key features and performance metrics of the
following use-cases: Immersive VR (IVR), teleoperation, automotive, Internet of drones,
interpersonal communications, live haptic-enabled broadcast, and cooperative automotive
driving.

TACTILE INTERNET TRENDSETTER FOR FUTURE GENERATIONS

URLLC IN TI APPLICATIONS

In URLLC systems, balancing the tradeoff between energy efficiency and latency is another key
consideration. Towards this purpose, an interesting future research direction is to analyses the
energy efficiency and delay tradeoffs in the unlicensed spectrum (i.e., 60 GHz) because the
regulations in the unlicensed band limit the maximum transmission time due to the requirement
for continuous channel sensing [3].
Another key research challenge is how to enable TI systems operating under strict latency and
reliability criteria to coexist with other systems such as eMBB systems demanding high
throughput applications and mMTC systems whose main objective is to support a massive
number of devices. How to efficiently multiplex URLLC traffic from TI systems and other
systems such as URLLC and eMBB in the downlink and uplink of future wireless systems is an
important research issue. Identifying TI traffic with guarantees across the access and network
layers is another important challenge [3].

LATENCY REDUCTION IN WIRELESS TI SYSTEMS

It is crucial to investigate novel methods across different segments of a wireless network to

further reduce the latency towards supporting TI applications in B5G networks. In the RAN
segment, millimeter wave communications seems to be a promising technology, several issues
including non-line-of-sight beamforming, Doppler, multi-path, atmospheric
absorpulence/attenuation need further attenuations [3].
DYNAMIC RESOURCE ALLOCATION IN TI SYSTEMS
Resources including communication resources, computing resources and caching resources are
distributed throughout the B5G wireless networks supporting TI applications in this direction,
future research works should focus on investigating suitable bandwidth-based and/or resource-
based slicing techniques, isolation techniques to isolate resources across user devices or
applications, and priority-based resource allocation and scheduling techniques for haptic
users/applications [3].
D. HAPTIC COMMUNICATIONS
There exist several challenges in employing haptic communications in future B5G networks.
Some of these challenges include the efficient design of haptic devices and haptic codecs,
maintaining the stability for control loop system. Due to time-varying delays and packet losses,
the system instability could be problematic in practical haptic communications systems [3].

E. WIRELESS AUGMENTED/VIRTUAL REALITY

The development of wireless AR/VR systems faces a number of challenges, including how to
manage both the tracking and QoS over highly dynamic wireless channels and how to design
high data-rate and low-latency links to the VR services. One of the main issues in wireless
AR/VR is to investigate suitable B5G network architecture. In contrast to the conventional VR
systems, which are mostly deployed over wired systems, wireless medium can provide much
higher freedom of movement of immersion. Wireless medium can provide much higher freedom
of movement of immersion. Deployment of VR applications in a wireless environment faces a
number of challenges. How to characterize the QoS of the VR systems, how to manage both the
tracking and QoS over highly dynamic wireless channels [3].

HARDWARE, COMPUTATIONAL AND NETWORK ARCHITECTURAL ISSUES

Future B5G networks should be designed to be flexible enough to support multiple vertical
applications such as haptic, M2M, V2V, and conventional video/data/voice while addressing
their different QoS requirements. However in practice, the network domain may comprise
different wireless networks, and this demands for suitable SDN and virtualization-based network
architectures to effectively coordinate among these networks [3].
In order to build effective tactile sensing systems, it is necessary to take into account the
sensitivity and ability to monitor measure many parameters of the sensor hardware, along with
the arrangement of tactile sensors and the physical challenges of spatial resolution and
conformability. In addition to the development of tactile sensors, it is essential to develop
hardware-friendly, low-complexity methodologies for processing the raw data captured by these
sensors [3].
 REFERENCES

[1] M. Gupta, R. K. Jha, and S. Jain, “Tactile based Intelligence Touch Technology in IoT configured
WCN in B5G/6G-A Survey,” IEEE Access, 2022, doi: 10.1109/ACCESS.2022.3148473.
[2] M. Simsek, A. Aijaz, M. Dohler, J. Sachs, and G. Fettweis, “5G-Enabled Tactile Internet,” IEEE
Journal on Selected Areas in Communications, vol. 34, no. 3, pp. 460–473, Mar. 2016, doi:
10.1109/JSAC.2016.2525398.
[3] S. K. Sharma, I. Woungang, A. Anpalagan, and S. Chatzinotas, “Toward Tactile Internet in
beyond 5G Era: Recent Advances, Current Issues, and Future Directions,” IEEE Access, vol. 8, pp.
56948–56991, 2020, doi: 10.1109/ACCESS.2020.2980369.
[4] D. Ohmann, M. Simsek, and G. P. Fettweis, “Achieving high availability in wireless networks by
an optimal number of Rayleigh-fading links,” in 2014 IEEE Globecom Workshops, GC Wkshps
2014, Mar. 2014, pp. 1402–1407. doi: 10.1109/GLOCOMW.2014.7063630.
[5] Institute of Electrical and Electronics Engineers, IEEE Computer Society, Institute of Electrical
and Electronics Engineers. Victorian Section, and IEEE Computer Society. Technical Committee
on Simulation, 2017 27th International Telecommunication Networks and Applications
Conference (ITNAC) : 22-24 Nov. 2017.
[6] C. She and C. Yang, “Ensuring the quality-of-service of tactile internet,” in IEEE Vehicular
Technology Conference, Jul. 2016, vol. 2016-July. doi: 10.1109/VTCSpring.2016.7504239.