PaperlayerView exact revision on arXiv

Source-linked AI summary

Survey on Terahertz Nanocommunication and Networking: A Top-Down Perspective

Filip Lemic, Sergi Abadal, Wouter Tavernier, Pieter Stroobant, Didier Colle, Eduard Alarcón, Johann Marquez-Barja, Jeroen Famaey

arXiv:1909.05703v3cs.NIcs.ETcs.SYeess.SY

TL;DR

Nanodevices need communication, but their constraints make conventional approaches difficult to apply. This survey synthesizes THz nanocommunication and nanonetworking research across applications, requirements, protocols, models, tools, and open challenges. It concludes by outlining current strengths and weaknesses and potential directions for future research.

  • Problem

    Supporting nanonetwork applications requires accounting for both application requirements and the capabilities of the nanodevices used.

    Source: A. Common Design Frameworks · p. 27

  • Method

    The survey takes a top-down view of THz nanonetwork applications and requirements, protocol layers, channel models, experimentation tools, and open challenges.

    Source: §I · p. 2

  • Results

    The survey outlines application requirements, current protocols, channel models, experimentation tools, their strengths and weaknesses, and potential future research directions.

    Source: §X · p. 31

  • Takeaways & Limitations

    Common design frameworks should account for both application requirements and the capabilities of the nanodevices used in each context.

    Source: A. Common Design Frameworks · p. 27

Abstract

from arXiv · show

Recent developments in nanotechnology herald nanometer-sized devices expected to bring light to a number of groundbreaking applications. Communication with and among nanodevices will be needed for unlocking the full potential of such applications. As the traditional communication approaches cannot be directly applied in nanocommunication, several alternative paradigms have emerged. Among them, electromagnetic nanocommunication in the terahertz (THz) frequency band is particularly promising, mainly due to the breakthrough of novel materials such as graphene. For this reason, numerous research efforts are nowadays targeting THz band nanocommunication and consequently nanonetworking. As it is expected that these trends will continue in the future, we see it beneficial to summarize the current status in these research domains. In this survey, we therefore aim to provide an overview of the current THz nanocommunication and nanonetworking research. Specifically, we discuss the applications envisioned to be supported by nanonetworks operating in the THz band, together with the requirements such applications pose on the underlying nanonetworks. Subsequently, we provide an overview of the current contributions on the different layers of the protocol stack, as well as the available channel models and experimentation tools. As the final contribution, we identify a number of open research challenges and outline several potential future research directions.

I. INTRODUCTION

Classical communication approaches do not readily fit nanoscale devices, while graphene-enabled THz electromagnetic communication offers a promising alternative. This survey reviews the field’s applications, requirements, protocols, models, tools, and research gaps.

I. INTRODUCTION

  • Nanoscale size, constrained power, uncertain material properties, and technological limits make classical electromagnetic approaches difficult to apply directly.A one-micrometer dipole would resonate at about 150 THz, where propagation loss also increases.Source: §I · pp. 1–2
  • Graphene supports THz surface plasmon polariton waves, enabling miniaturization, lower-frequency operation, and voltage-tunable antennas for nanocommunication.The survey focuses on electromagnetic communication, which the authors identify as one of the most promising paradigms and suitable for several propagation media.Source: §I · pp. 1–2
  • The survey connects THz nanonetwork applications and their requirements with protocol-layer research, channel models, experimentation tools, and open challenges.Its top-down approach is intended to map application requirements to protocols and their design goals.Source: §I · p. 2

A. Methodology

  • The literature review uses a systematic process: title and abstract screening in Google Scholar, followed by full-text eligibility assessment.It includes English, peer-reviewed full novel solutions and places extra emphasis on work published since 2015, highly cited work, and major venues.Source: A. Methodology · p. 2

B. Structure

  • The paper proceeds from related surveys and application requirements through protocol layers, channel models, experimentation tools, and general open challenges.The protocol-layer reviews cover network, link, and physical layers.Source: B. Structure · p. 2

II. RELATED WORK

  • Earlier surveys often predate recent THz nanonetworking work or cover only general THz communication, nanocommunication, IoNT, or MAC protocols.This survey focuses exclusively on electromagnetic THz nanocommunication and nanonetworking; it excludes THz macro-scale topics, other nanocommunication paradigms, and hardware.Source: §II · p. 3

III. APPLICATION DOMAINS ENABLED BY THZ NANONETWORKS

The survey considers application domains for THz nanonetworks and the requirements they place on supporting networks. For software-defined metamaterials, those requirements vary by application stage and update rate.

III. APPLICATION DOMAINS ENABLED BY THZ NANONETWORKS

  • Application requirements are summarized by domain, using the most stringent requirements while recognizing that some applications may be less demanding.The survey distinguishes domains such as wireless robotic materials and body-centric communication by their requirements, even when both include sensing-only applications.Source: §III · p. 3

A. Software-Defined Metamaterials

  • Software-defined metamaterials reconfigure subwavelength unit cells through software instructions, addressing the limitation of cells hard-coded for one application or condition.Proposed uses include programmable wireless environments, simplified wireless transmitters, and dynamic holographic imaging.Source: A. Software-Defined Metamaterials · p. 4
  • THz communication is proposed for an intra-metamaterial network connecting controllers, because unit-cell scale imposes a frequency-dependent form-factor limit.Envisioned metamaterials may contain thousands to millions of controllers, with low energy consumption and energy harvesting rather than batteries.Source: A. Software-Defined Metamaterials · pp. 4–5
  • Early software-defined metamaterial prototypes are expected to use static nodes, mostly downlink control traffic, relaxed latency, and low reliability and security requirements.These assumptions support proof-of-concept prototypes with simple controllers and network infrastructure.Source: A. Software-Defined Metamaterials · p. 5
  • The intra-metamaterial network’s latency should be a fraction of the application’s refresh time, while its throughput estimate depends on assumed unit-cell count, area, range, and update pattern.A simplified 10-by-10-controller, 100 ms flooding scenario yields 14.4 kbps; the authors describe the resulting 1 to 50 kbps rule of thumb as a rough minimum indication.Source: A. Software-Defined Metamaterials · p. 5
  • For second-generation software-defined metamaterials, estimated minimum throughput is 144 kbps at a 10 ms update period and 144 Mbps at 10 µs.The broader rule-of-thumb range is 50 kbps to 500 Mbps; the derivation assumes each controller transmission also serves as a command to its metamaterial.Source: A. Software-Defined Metamaterials · p. 6

B. Wireless Robotic Materials

The survey describes wireless robotic materials, in-body communication, and on-chip communication as THz nanonetwork applications with distinct network requirements. It relates those requirements to application scale and traffic, and presents THz electromagnetic nanocommunication as a candidate across the discussed domains.

B. Wireless Robotic Materials

  • Wireless robotic materials integrate sensing, actuation, computation, and communication in programmable composites, with applications ranging from adaptive airfoils to tactile skins.The survey expects dense swarms and estimates that a patch of 100 elements updated every 20 ms would require roughly 100 kbps to 10 Mbps, with bidirectional traffic intuitively needed for sensing and actuation.Source: B. Wireless Robotic Materials · pp. 6–7
  • For wireless robotic materials, some applications require high reliability, security, mobility, and localization, while energy demands vary with device size and lifetime.The survey distinguishes these sensing-and-actuation applications from sensing-only wireless nanosensor networks by treating the latter as a subset.Source: B. Wireless Robotic Materials · pp. 6–7

C. In-body Communication

  • In-body nanodevices are envisioned to sense or act within the body while communicating with the outside world, with network structure and traffic depending on the application.For a modeled brain-sensing swarm of 10^6 devices, the survey estimates 24 Mbps and gives a rule-of-thumb throughput range of 1 to 50 Mbps; these requirements are derived for the most challenging applications.Source: C. In-body Communication · pp. 7–8
  • Body-centric applications may require very low energy use, high reliability, low end-to-end latency, strong security, and support for mobile, potentially localizable nanodevices.The passages associate low latency with real-time control and note that mobility can arise from blood flow or body movement.Source: C. In-body Communication · pp. 7–8

D. On-chip Communication

  • On-chip communication became a performance and energy concern as multicore chips added cores, while wired NoCs faced delay, power, and area challenges.The survey describes wireless on-chip communication as an alternative interconnect technology.Source: D. On-chip Communication · p. 8
  • Graphene nanoantennas a few micrometers in size could support tunable THz inter-core links, while on-chip networks face stringent density, latency, throughput, and reliability constraints.On-chip topology and propagation characteristics can be considered static and known upfront; reported throughput requirements are 10–100 Gbps and can reach the Tbps barrier in communication-intensive architectures.Source: D. On-chip Communication · pp. 8–9

IV. NETWORK LAYER

The network layer must support communication over arbitrary distances despite nanonode hardware, memory, channel, and energy limits. The survey organizes work around forwarding and relaying, flooding, and topology-specific routing, including approaches that account for THz channel effects.

IV. NETWORK LAYER

  • Network-layer routing must work around limits on control-message exchange, memory, channel access, energy, and scale, while some THz protocols incorporate molecular absorption into path selection.The passage contrasts application scales that may exceed the up-to-200-mote range for which many WSN routing protocols are optimized.Source: §IV · p. 9
  • The survey classifies network-layer work into forwarding and relaying, flooding-based one-to-many routing, and one-to-one routing for regular or irregular topologies.Forwarding studies consider direct transmission to a farther hop versus routing through one or more intermediaries.Source: §IV · p. 10

A. Relaying and Forwarding for THz Nanonetworks

  • THz relaying and forwarding studies report gains over direct transmission, including 2.2 dB for amplify-and-forward and 5 dB for decode-and-forward, and 12–13× end-to-end capacity with delay increase below 0.02%.The channel-aware forwarding scheme avoids relays in regions adversely affected by molecular absorption and avoids short hops that increase hop count.Source: A. Relaying and Forwarding for THz… · p. 10
  • A hierarchical cluster architecture halves energy per bit for forwarding versus direct transmission at 1–10 mm, while improving delay by a factor of 3–4.A separate event-reporting policy based on a Markov Decision Process exceeds comparison policies in throughput by at least a couple of percent, with a larger advantage as events become more frequent.Source: A. Relaying and Forwarding for THz… · pp. 10–11

B. Flooding-based Routing

Flooding-based routing reduces redundant transmissions by limiting or selecting forwarders, but its efficiency and reliability depend on communication pattern and network setting. The supplied passages also describe trade-offs among coverage, memory, latency, and routing constraints.

B. Flooding-based Routing

  • Adaptive and selective flooding reduce redundant transmissions by limiting forwarders or choosing them probabilistically.Liaskos et al. report approximately 10% improvement in packet rate to achieve full coverage, with similar latency and removal of interference effects.Source: B. Flooding-based Routing · p. 11
  • ProbA achieves full coverage in the aorta with 40% of nanonodes participating, but fails to achieve full coverage in the narrower artery.Fixed-probability flooding requires approximately 60% participation in both environments.Source: B. Flooding-based Routing · p. 11
  • LENWB achieves full coverage with 50% participation in the aorta and 20% in the artery, at a significant memory cost.In the IoNT approach, energy-constrained next-hop selection extends network lifetime at the cost of 2–3 times higher latency than shortest-path collection.Source: B. Flooding-based Routing · p. 11
  • Flooding is inefficient for one-to-one delivery, while reducing transmissions can sacrifice delivery reliability or suitability for mobile nodes.Probabilistic approaches may fail if poorly tuned, and topology-dependent approaches are less useful when nodes move.Source: B. Flooding-based Routing · p. 11

D. Irregular Topologies

Routing approaches for irregular topologies use local coordinates or fault information, but static assumptions and incomplete evaluation limit their demonstrated scope. The survey also highlights gaps in energy measurement, scalability, mobility, and comparability.

D. Irregular Topologies

  • Tsioliaridou et al. use fixed anchors and triangulation for non-unique coordinates, then selectively flood toward destinations without per-packet state.The approach was compared with two flooding approaches, which required five times more nanonodes to decode messages in the reported paper; this depends on network size.Source: D. Irregular Topologies · p. 13
  • Coordinate-based routing schemes offer one-to-one routing but require static environments, since mobility frequently invalidates coordinate systems.The works also do not address how nodes obtain other nodes’ coordinates or addresses.Source: D. Irregular Topologies · p. 13
  • Energy evaluations based only on message counts are incomplete because per-message power also depends on transmission distance and message length.Source: D. Irregular Topologies · p. 13
  • Routing scalability is often unevaluated, and studies that vary network size find coverage and power consumption depend non-linearly on area or volume and node density.Fixed-size evaluations may therefore yield conclusions that do not generalize to larger or smaller networks.Source: D. Irregular Topologies · p. 13
  • In-body mobility can impose costly routing-information updates, while varied topologies and metrics make results difficult to compare.The survey notes that numerical comparisons across settings, such as Stelzner et al., are rare.Source: D. Irregular Topologies · p. 14

V. LINK LAYER

THz nanonetwork link layers coordinate channel access and recover from bit errors, but classical protocols face the band’s wide bandwidth, complexity, and energy-harvesting constraints. The survey reviews distributed protocols, including schemes that coordinate transmission with energy availability or receiver feedback.

V. LINK LAYER

  • Classical link-layer protocols are not directly applicable because THz bandwidth is nearly 10 THz, carrier-sensing can be too complex, and harvested energy changes device availability.Source: §V · p. 14

A. Distributed Protocols

  • Akkari et al. schedule local transmissions to support the largest set of traffic rates while meeting hard delivery deadlines and accounting for energy harvesting.Their decisions use a CSMA Markov chain and Lyapunov optimization based on incoming traffic and queue length.Source: A. Distributed Protocols · pp. 14–15
  • PHLAME uses a handshake for transmitter and receiver to jointly select symbol rate and coding, and is reported to support dense networks across energy per useful bit, delay, and throughput.Source: A. Distributed Protocols · p. 15
  • DRIH-MAC initiates communication with receiver Ready-to-Receive packets and uses probabilistic edge-coloring-based scheduling.The authors claim scalability, low weight, reduced collision probability, and increased harvested-energy utilization.Source: A. Distributed Protocols · p. 15
  • Other distributed designs include timing-channel communication with acknowledgments for error recovery, and receiver-initiated flow control that limits reception to the receiver’s current energy level.TCN encodes information in silence duration; the synchronization protocol in uses a sliding flow-control window.Source: A. Distributed Protocols · p. 15

B. Hierarchical Protocols

Hierarchical link-layer protocols shift coordination to more capable nodes, while the survey also reviews on-chip MAC designs from sub-THz work. It identifies unresolved link reliability, latency, mobility, and evaluation challenges.

B. Hierarchical Protocols

  • The hierarchical MAC in uses nanocontroller-managed TDMA and symbol-compression scheduling to optimize throughput while keeping transmission ratios below the Critical Transmission Ratio.Elastic inter-symbol spacing permits parallel transmissions without collisions.Source: B. Hierarchical Protocols · pp. 15–16
  • EEWNSN-MAC clusters mobile nodes around static nanorouters, then uses TDMA scheduling before aggregation and forwarding to a nano-micro interface.Source: B. Hierarchical Protocols · p. 16

C. Protocols for On-Chip Communication

  • The on-chip protocols reviewed use sub-THz rather than THz frequencies, and their applicability to THz remains unevaluated.THz signals attenuate faster and resonate with water, unlike mmWave signals, which resonate with oxygen molecules.Source: C. Protocols for On-Chip Communication · pp. 16–17
  • Most existing protocols do not optimize delivery latency, and protocols jointly optimizing throughput and latency for on-chip communication are missing.Source: C. Protocols for On-Chip Communication · p. 17
  • Existing protocol evaluations rely on analysis or simulation, may omit idle energy, and use non-standardized, non-exhaustive metrics that prevent comprehensive comparison.Scalability is evaluated only for EEWNSN-MAC, whose conclusion is that it is potentially scalable.Source: C. Protocols for On-Chip Communication · p. 18

VI. PHYSICAL LAYER

THz nanocommunication imposes tight area, energy, and memory constraints, while immature THz technology and distance-dependent molecular absorption further challenge the physical layer. Existing designs favor simple modulation and coding: CW OOK suits sustained-power on-chip links, whereas impulse-radio approaches target intermittent energy availability.

VI. PHYSICAL LAYER

  • Nanoscale resource limits, immature THz circuits, and distance-dependent molecular absorption constrain physical-layer design.The constraints include area, energy, and memory, and can affect reliability when energy harvesting makes device operation intermittent.Source: §VI · p. 18

A. Continuous-Wave (CW) vs. Impulse Radio (IR)

  • CW transceivers from 0.06 to 0.43 THz reach 10+ Gb/s at around 1 pJ/bit for high-rate on-chip, off-chip, and indoor applications.The survey expects the trend to continue as efforts address the THz gap.Source: A. Continuous-Wave (CW) vs. Impulse… · p. 18, Figure 9 · p. 19
  • On-chip proposals favor simple OOK and non-coherent detection, avoiding power-hungry coherent circuits under low-latency and low-power requirements.OOK is 1.5X and 2.5X more energy efficient than BPSK and QPSK, respectively, in on-chip environments; scaling rates may require multiple carriers to combat dispersion.Source: A. Continuous-Wave (CW) vs. Impulse… · pp. 18–19
  • For intermittent energy availability, carrier-less pulse-based BPSK is the most efficient and reliable of the compared IoNT modulations, despite more complex decoding.OOK and PPM are simpler but less reliable and efficient; the analysis discourages PAM for low performance and efficiency.Source: A. Continuous-Wave (CW) vs. Impulse… · p. 19

B. Pulse-based Modulations

Pulse-based modulation research adapts TS-OOK to multi-user access, energy efficiency, and THz channel conditions. Pulse-based coding seeks to reduce energy and interference while maintaining reliability, including receiver-aware error control for energy-harvesting settings.

B. Pulse-based Modulations

  • TS-OOK uses roughly 100-fs pulses separated by much longer silences, retaining OOK simplicity for energy-limited or energy-harvesting applications.Its original proposal reports capacity on the order of Tbps and introduces an interference model later validated experimentally.Source: B. Pulse-based Modulations · p. 19
  • RDMA assigns co-prime user rates to reduce interference, while pseudo-random time-hopping variants address rate imbalance or adapt rates to user needs.Direct-sequence codes offer another multi-user approach, sacrificing energy efficiency for higher performance.Source: B. Pulse-based Modulations · pp. 19–20
  • Combining TS-OOK with PPM can encode multiple bits in the silence between pulses, improving energy efficiency at the cost of data rate.A time-hopping PPM variant was evaluated with non-coherent detection and multiple modulation orders.Source: B. Pulse-based Modulations · p. 20
  • THz-specific proposals use frequency hopping to handle composition-varying absorption without channel-state observation, or distance estimation for static channels to select waveforms or frequencies.Graphene-based beam hopping is also proposed to switch spatial channels and further reduce interference in TS-OOK scenarios.Source: B. Pulse-based Modulations · p. 20

C. Pulse-based Coding

  • Low-weight coding uses silences to save power and mitigate interference without reducing each user's transmission rate; later work derives energy-minimizing codebooks under broader constraints.Prefix-free codes support instantaneous decoding, while simple block nanocodes add reliability at very small cost; a comparative study evaluates energy efficiency, bandwidth expansion, robustness, and interference.Source: C. Pulse-based Coding · p. 20
  • A probing-packet error-control strategy for energy harvesting sends data only after receivers acknowledge that they expect sufficient energy to receive it.Receivers remain silent when their energy is low.Source: C. Pulse-based Coding · p. 20

D. Beaming and Detection

The survey reviews simplified beaming and detection methods for resource-constrained THz nanocommunication, alongside the hardware and synchronization challenges that remain. It identifies compact, efficient THz signal generation—especially for very short pulses—as a major implementation hurdle.

D. Beaming and Detection

  • Graphene antennas can be miniaturized and rapidly reconfigured by changing electrostatic voltage, enabling simple control of beam direction and resonance frequency.A programmable PHY interface exposes beam switching and frequency tunability to upper layers, including bit-level beam switching for TS-OOK beam multiplexing.Source: D. Beaming and Detection · p. 20
  • CTMA detection with a low-pass filter and peak detector has demonstrated potential for Tbps detection with relaxed synchronization requirements.Related proposals use delayed CTMA detectors for joint detection and synchronization in TS-OOK, or time-of-arrival estimation for time-hopping PPM.Source: D. Beaming and Detection · p. 20

DS-OOK [176]

  • Compact, efficient THz signal generation remains a major implementation challenge, particularly for the hundred-femtosecond pulses assumed in most proposals.The physical layer is well researched theoretically and in simulation, but the analog transceiver front-end remains a hurdle to realization.Source: DS-OOK [176] · p. 21
  • Synchronization remains a central impulse-radio challenge, and potentially very high sampling speeds for Tbps operation conflict with simplicity and efficiency demands.Non-coherent detectors and CTMA approaches relax synchronization requirements but do not remove the challenge.Source: DS-OOK [176] · p. 21

VII. CHANNEL MODELING

THz channel models must capture propagation losses, dispersion, and noise, with molecular absorption and scattering especially relevant at these wavelengths. The surveyed models account for frequency-selective absorption and noise, scattering, rough-surface effects, and application-specific propagation conditions.

VII. CHANNEL MODELING

  • THz channel modeling must account for molecular absorption, which creates frequency-selective attenuation and added noise, as well as scattering and losses from rough reflections.These impairments can restrict bandwidth, and the relevant channel depends on the application scenario.Source: §VII · pp. 21–22

A. THz Propagation Models

  • A propagation model combines spreading loss, molecular absorption, and particle scattering; rough-surface reflections may require an additional scattering term.The model also includes system noise, molecular noise, and other noise sources.Source: A. THz Propagation Models · p. 22
  • Molecular absorption attenuation depends on frequency, medium composition, and transmission distance, and can limit channel bandwidth.The cited model also represents molecular noise; studies report absorption and dispersive effects are generally negligible up to a few centimeters in the short-distance setting considered.Source: A. THz Propagation Models · p. 22, Figure 11 · p. 22
  • Particle and rough-surface scattering are frequency-selective impairments that may be non-negligible at certain distances.Particle-scattering models require particle density and scattering cross sections; measurement campaigns characterize rough-surface responses across materials including wood, plaster, concrete, plastic, glass, and metal.Source: A. THz Propagation Models · pp. 22–23
  • Ray-based methods support THz channel characterization, including interference and SINR or spectral-efficiency evaluation for multiplexed links.In the cited multi-user scenarios, interference becomes more critical than molecular absorption, and the work estimates receiver spacing to maximize area capacity.Source: A. THz Propagation Models · p. 23

B. Intra-body Channels

Intra-body THz propagation depends on tissue composition, transmitter and receiver location, and changing biological conditions. Channel and noise models have expanded to account for these factors, but their effects remain substantial.

B. Intra-body Channels

  • Tissue-specific dielectric loss and transitions between tissues affect spreading loss, while tissue composition and particles affect molecular absorption and scattering.These factors vary with device position and patient, and the model terms affected are Aspr, Amol, and Asca.Source: B. Intra-body Channels · p. 23
  • Around 30 dB/mm attenuation was reported for fat; a multilayer body model found up to 30% of incident external power may be reflected.The reflection result was reported as symmetrical; other work modeled blood, skin, and fat to derive noise and channel capacity.Source: B. Intra-body Channels · p. 23
  • Noise models progressed from assuming molecular-absorption noise dominates to including body radiation, transceiver thermal noise, Doppler effects, and interference.Comparisons validated molecular-absorption noise dominance; combined noise and interference models derive SINR metrics for capacity and throughput.Source: B. Intra-body Channels · p. 23
  • Intra-body channels can vary over time with temperature, pressure, or molecular composition, including blood composition across breathing phases.Respiration also changes lung volume and composition, and vegetation attenuation can vary over time with photosynthesis.Source: B. Intra-body Channels · p. 23

C. On-chip Communication Channels

Realistic THz on-chip channel models remain limited: unpackaged free-space assumptions miss package effects, and silicon can impose substantial path loss. Simulation tools provide varied capabilities, but their realism, scalability, and experimental validation remain open concerns.

C. On-chip Communication Channels

  • A realistic chip channel must capture package losses and multipath; controlled geometry makes it nearly time-invariant and quasi-deterministic.Through-silicon vias and dielectric transitions matter, while molecular absorption and diffuse scattering are not problems in this setting.Source: C. On-chip Communication Channels · pp. 23–24
  • On-chip wireless propagation measurements up to 300 GHz show metallic enclosures can act as reverberation chambers, with low path loss but very high delay spread.This evidence comes from small enclosed environments, not nanoscale chip packages.Source: C. On-chip Communication Channels · p. 24
  • Existing THz chip-channel work includes ray tracing in stratified structures, but much prior work assumes unpackaged free-space propagation.Sub-THz studies extend to around 140 and 200 GHz; the cited THz model extracts path loss and dispersion metrics.Source: C. On-chip Communication Channels · p. 24
  • Realistic on-chip THz channel models remain scarce, and silicon losses remain unresolved; realistic packages can add tens of decibels of path loss.At THz frequencies, the losses may worsen because antenna effective aperture becomes smaller.Source: C. On-chip Communication Channels · pp. 24–25

VIII. SIMULATION AND EXPERIMENTATION TOOLS

  • NanoSim, Vouivre, BitSimulator, and TeraSim offer different THz simulation capabilities, spanning channel, physical-layer, mobility, and network functions.NanoSim supports configurable network layers; Vouivre and BitSimulator use TS-OOK, while TeraSim includes energy harvesting and nanoscale link protocols.Source: §VIII · pp. 25–26
  • TeraSim appears most extensive, while BitSimulator trades complexity for scalability; the scalability-versus-realism trade-off remains untested.The simulators' scalability and the reliability of their differing results still need evaluation.Source: §VIII · pp. 25–26
  • Energy-harvesting simulations remain simplified, and accurate nanonode energy models should include nonlinear charging, finite storage, idling, processing, and storage costs.Human mobility models and their integration with simulators are also needed; experimental facilities and public datasets remain largely uncharted.Source: §VIII · p. 26

IX. ADDITIONAL CHALLENGES, OPEN ISSUES, AND FUTURE RESEARCH DIRECTIONS

The survey identifies protocol objectives as a way to match designs to application requirements and expose gaps. It also calls for common frameworks that jointly account for application demands and device capabilities.

IX. ADDITIONAL CHALLENGES, OPEN ISSUES, AND FUTURE RESEARCH DIRECTIONS

  • Table VII organizes existing protocols by optimization objective, helping readers select protocols for application requirements and identify improvement directions.The objectives are design aims, distinct from the performance metrics used to evaluate protocols.Source: §IX · p. 26, Table VII · p. 27

A. Common Design Frameworks

  • A common design framework should jointly consider each application's nanonetwork requirements and the capabilities of the devices used in that setting.The same device types are unlikely to suit on-chip, in-body, and software-defined metamaterial applications alike.Source: A. Common Design Frameworks · p. 27
  • Application-level requirements remain sparsely specified, with existing work mostly focused on in-body communication; exhaustive qualitative requirements are needed across domains.Device specifications should also cover size, available energy, processing power, and memory.Source: A. Common Design Frameworks · p. 27

B. Transport Layer Protocols

Transport protocols and energy lifecycle models remain unsettled for constrained nanonetworks. The survey describes reduced-stack approaches and energy models while highlighting narrow protocol coverage and assumptions that can undermine feasibility.

B. Transport Layer Protocols

  • Transport-layer proposals address software-defined metamaterials and body-centric networks, but their performance details and coverage of other applications remain limited.Scalability and protocol overhead are unclear, and extra signaling may consume nanonode energy.Source: B. Transport Layer Protocols · p. 28
  • Whether transport protocols are needed in many scenarios is unclear, since their overhead may be infeasible for energy-constrained nanonodes.The surveyed transport work is still largely unexplored beyond the two narrow application domains discussed.Source: B. Transport Layer Protocols · p. 28

C. Reduced / Integrated Protocol Stack

  • Energy-harvesting nanonodes motivate condensed, application-specific protocol stacks, but fuller stacks may increase energy use, latency, and complexity.The trade-off between reducing stack complexity and providing a full stack remains largely unexplored.Source: C. Reduced / Integrated Protocol Stack · p. 28
  • BitSurfing interprets symbols from an external source and emits a short, low-energy pulse when a desired message appears, without link or transport layers.The authors report perpetual operation; the cited harvesting estimate is about one packet transmission per 10 seconds.Source: C. Reduced / Integrated Protocol Stack · p. 28

D. Energy Lifecycle Modeling

  • Energy lifecycle models connect harvesting and TS-OOK bit transmission or reception costs to packet delivery probability, end-to-end delay, and throughput.Related models also examine packet size, repetitions, and code weights.Source: D. Energy Lifecycle Modeling · p. 29
  • A micrometer-sized piezoelectric system in lossy environments becomes inoperative beyond 1.5 mm; feasible communication requires idling energy nine orders below reception energy.Current nanoscale systems achieve, at best, idling energy only three orders below reception energy.Source: D. Energy Lifecycle Modeling · p. 29
  • Models that assign all device energy to transmission and reception omit wake-up and other functions, so protocols may be infeasible under more realistic energy accounting.Energy models should also account for sensing, processing, and energy-harvesting and consumption lifecycle effects.Source: D. Energy Lifecycle Modeling · p. 29

E. End-to-End Architectures

The survey reviews end-to-end integration needs and identifies gaps in security, localization, and standardization for THz nanonetworks. It also consolidates the state of the art and future research directions.

E. End-to-End Architectures

  • Seamless nanonetwork integration requires gateways between macro- and nano-worlds, while current work largely targets uplink rather than control-oriented downlink.One proposed backhaul mechanism uses on-demand polling by nanosinks and accounts for dynamic bandwidth and THz channel conditions.Source: E. End-to-End Architectures · pp. 29–30

F. Security, Integrity, and Privacy

  • THz nanonetwork security remains largely unexplored, with open needs for key management, low-overhead protection, and attack detection and response.The stated goals are confidentiality, integrity, and availability.Source: F. Security, Integrity, and Privacy · p. 30

G. Localization and Tracking

  • Localization must balance high accuracy and low energy use; multi-hop errors accumulate, while backscatter-based ranging still faces propagation and hardware challenges.Existing hop-counting and pulse-based methods can fit energy constraints but are expected to have relatively low accuracy; a backscatter approach estimates round-trip time-of-flight from multiple anchors.Source: G. Localization and Tracking · p. 30

H. Standardization

  • IEEE P1906.1 offers a conceptual framework, but omits device and transmission parameters, OSI layers 2–3 techniques, and recommendations for network interconnection.The survey identifies further standardization needs including media access control, addressing, flow control, error detection, and routing.Source: H. Standardization · p. 30

X. CONCLUSION

  • The survey synthesizes application domains, their network requirements, current protocols, channel models, and experimentation tools, then outlines future research directions.Proposed directions include prototypes and experimental infrastructures, higher-layer protocols, mobility mitigation, security, and localization.Source: §X · p. 31

Related papers

Loading 1909.05703v3…