On-Demand Routing for Urban VANETs using Cooperating UAVs

Abderrahmane Lakas; Noureddine Chaib; Omar Sami Oubbati; Salim Bitam

REVIEW 4 major objections 5 minor 13 references

On-Demand Routing for Urban VANETs using Cooperating UAVs

T0 review · 4 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A new reactive routing protocol that makes UAVs full members of discovered paths and uses them as bridges for fragmented ground segments can extend connectivity in sparse urban VANETs, reducing end-to-end delay and raising packet delivery…

desk verdict Plausible protocol, unenforced UAV-coverage assumption, and a conclusion that contradicts its own delay plot. read the letter →

arxiv 1908.05905 v1 pith:NTB7RK7E submitted 2019-08-16 cs.NI cs.ITcs.SImath.IT

Omar Sami Oubbati , Noureddine Chaib , Abderrahmane Lakas , Salim Bitam This is my paper · ORCID

classification cs.NIcs.ITcs.SImath.IT

keywords VANETroutingUAV-assistedreactiveurbanvehicularnetworkspathscoringnetworkconnectivitypacketdeliveryratioend-to-enddelay

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to establish that a reactive routing protocol for urban vehicular networks can keep delivery reliable in sparse traffic by recruiting drones as full members of routing paths and as bridges when ground connectivity fragments. The proposed scheme scores every discovered path by vehicle count, delay, density balance, and number of UAV hops, then selects the best path and keeps alternative paths for instant failover. In packet-level simulation against two existing urban VANET protocols, it reports higher packet delivery ratio and lower end-to-end delay, especially when the vehicle count is low. If true, the significance is that heterogeneous vehicle-drone communication is a practical way to extend a VANET's usable range without waiting for dense traffic.

What carries the argument

The central object is a routing path represented as a sequence of fixed zone IDs, where each zone is roughly the size of the vehicle communication range (about 300 m). During route discovery, a RREQ packet accumulates each transited zone ID and the vehicle count in that zone, so the destination can compute the average zone density and the standard deviation of zone densities for every discovered path. The selection score combines those computed metrics with the raw delay and the number of UAV hops, and the winning path is returned to the source greedily. The maintenance machinery stores all discovered paths in the data-packet header, so an intermediate node can switch to an alternative path without re-running discovery, and when no ground alternative exists, a UAV is designated as the next forwarder to bridge disconnected clusters.

What would settle it

Run the same protocol in a simulator where drone coverage is not guaranteed over every four road segments, or where building obstacles attenuate air-to-ground links, and measure whether the packet-delivery-ratio advantage over ground-only routing vanishes below a threshold such as fewer than the assumed 16 UAVs.

Watch

Extended reading notes

Core claim

The paper claims that a heterogeneous network in which UAVs are full members of on-demand routing paths, rather than mere data mules, keeps urban vehicular networks connected when ground traffic is sparse. The destination scores each candidate path by $$\mathrm{Score}=\frac{\mathrm{NB_{vehicles}}}{\mathrm{Delay}}\times \frac{1}{1+\mathrm{Sdeviation}+\mathrm{Hops_{UAV}}}$$ where NB_vehicles is the number of vehicles along the path, Delay is the RREQ traversal time, Sdeviation is the standard deviation of per-zone vehicle counts, and Hops_UAV is the number of UAV hops in the path. Paths with more vehicles, lower delay, balanced vehicle distribution, and fewer UAV hops are preferred; all discovered paths are stored in the data-packet header, and when the selected path breaks, the current forwarder switches to an alternative path, or calls on a UAV to bridge the gap. In NS-2 simulations over a 4 km x 4 km grid with up to 200 vehicles and 16 UAVs, the protocol reports higher packet delivery ratio and lower end-to-end delay than RBVT-R and AGP at low vehicle densities, though at high densities RBVT-R achieves lower delay.

Load-bearing premise

The scheme needs a sufficient number of UAVs that each area of four road segments always has at least one UAV overhead, and it assumes drone-to-vehicle wireless links are long-range and unaffected by buildings; without those two conditions the connectivity gain disappears.

Editorial extensions

If this is right

In sparse urban traffic, routing paths need not be confined to ground vehicles: UAV hops extend connectivity and reduce the average number of hops.
Storing all discovered paths in the data-packet header lets intermediate nodes switch paths without a new route discovery as long as at least one alternative remains.
Density balance, not just shortest distance or fewest intersections, is a workable route-selection criterion for city VANETs.
At high vehicle densities, ground-only reactive routing can match or beat the UAV-assisted approach in end-to-end delay, because UAV mobility triggers route rediscovery.
When alternatives are exhausted, the protocol falls back to a route-error message and a fresh discovery, so failures degrade to standard reactive behavior.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The protocol assumes blanket drone coverage over every group of four road segments; a more economical design would reposition UAVs toward predicted disconnection points rather than hovering everywhere.
The score weights vehicle density, delay, density balance, and UAV hops equally; changing those relative weights could trade path stability against delay and could be tuned separately for highway or rural settings.
If air-to-ground links suffer the same building blockage as ground-to-ground links, the bridge advantage shrinks, so a simulation varying building height and link budget would test the core connectivity claim.
The zone-density scoring idea could be reused in non-urban settings, which the paper itself names as a possible extension.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Colleague,

The thing to know: this paper has a sensible reactive routing protocol that combines zone-density scoring with UAV relays, but the simulation evidence is weaker than the abstract claims, and the load-bearing assumption about UAV coverage is never actually enforced.

What's new: the specific combination — fixed zones, standard-deviation-based density balance, alternative paths stored in the data header, and UAVs as a final bridge when the ground network fragments — doesn't appear in the cited prior work. That is a legitimate, incremental extension. The scoring equation is explicit and not fitted to the outcomes, so there's no circularity. The maintenance idea, using pre-discovered paths before falling back to drones, is a reasonable design choice.

The soft spots, in rough order. First, Section III.A assumes enough UAVs that one hovers over every four road segments, but the simulation uses 16 UAVs moving at 50–120 km/h with no mechanism that puts them at fragmentation points. That means the improved PDR might come from fortuitous random placement rather than the protocol's discovery and maintenance logic. The authors even list 'make the mobility of UAVs more controllable' as future work, which tells you they know the assumption isn't implemented. Second, the conclusion claims the protocol outperforms on delay 'especially the high densities,' which is directly contradicted by Fig. 7b and their own results section, where RBVT-R wins above 140 vehicles. That needs fixing — it's a simple overstatement, but in a short paper it undermines trust. Third, there are no error bars or statistical tests despite 15 repeats, and no code or data release. Fourth, the closest baseline is the authors' own 2017 UAV-assisted protocol, which is cited but not compared against.

The math is simple and mostly fine; the scoring equation is arbitrary but not derived by fitting, so no circularity. The citation pattern is clean.

This paper is for people working on UAV-assisted VANET routing. It deserves a serious referee only if the coverage assumption is addressed and the delay conclusion corrected. As is, I'd treat it as a conditional paper with a promising core. A journal editor could reasonably send it to review, because the flaws are addressable and the protocol idea is not silly — but I'd expect a major revision, not a quick accept.

Recommendation: engage with it, but make the authors prove the UAV placement and fix the internal contradiction before relying on the results.

Referee Report

4 major / 5 minor

Summary. The paper proposes a reactive routing protocol for urban VANETs in which UAVs participate in route discovery, path scoring, and path maintenance. Paths are scored using a multi-criteria formula that rewards vehicle count, penalizes delay, density imbalance, and the number of UAV hops, and the best path is selected from multiple discovered routes. When a route breaks, an alternative pre-discovered path is used, and UAVs are employed to bridge disconnected ground clusters. The protocol is evaluated in NS-2 with VanetMobiSim/MobiSim mobility against RBVT-R and AGP, reporting packet delivery ratio, end-to-end delay, hop count, and routing overhead. The abstract claims reduced end-to-end delay and high delivery ratio, supporting the conclusion that heterogeneous vehicle-UAV communication extends network connectivity.

Significance. If the claimed results hold, the paper would make a moderate contribution to UAV-assisted VANET routing by combining zone-density-aware path scoring with UAV relaying and pre-computed alternative paths. The comparison against RBVT-R and AGP is appropriate, and the scoring rule in Eq. (3) is a constructive, parameter-free heuristic rather than a fit to the reported outcomes. However, the central claim rests entirely on simulation evidence that conflicts with the paper's own coverage assumption and with its internal delay statements; no statistical support is provided. The idea is plausible and worth pursuing, but the current manuscript does not establish the load-bearing connectivity claim.

major comments (4)

[III.A, Table II] The coverage assumption in Section III.A that the network has a sufficient number of UAVs so that at each moment at least one UAV hovers over an area of four road segments is never enforced or verified in the simulation. Table II specifies only 16 UAVs moving at 50–120 km/h, and the conclusion lists making UAV mobility more controllable as future work. Since the claimed connectivity gain depends on a UAV being within communication range of two disconnected ground segments at the instant a path is needed, the simulation results may reflect random UAV placement rather than the routing protocol itself. The authors should either implement a UAV control/placement model that realizes the hovering assumption or evaluate the protocol under conditions where the assumption is relaxed.
[IV.C, Fig. 7(b), Conclusion] The results text in Section IV.C explicitly states that for more than 140 vehicles, RBVT-R outperforms the proposed approach in end-to-end delay, and that UAV-composed routes are unstable due to high mobility. The Conclusion, however, claims the protocol outperforms existing routing in terms of end-to-end delay 'especially the high densities,' and the abstract states an unqualified 'reduced end-to-end delay.' These statements are internally contradictory and overstate the findings. The claims must be restricted to the density range where the data support them.
[IV.A, IV.C] The evaluation reports 15 simulation repeats per scenario, but no error bars, confidence intervals, or statistical tests are presented for any metric. This is particularly important because the end-to-end-delay advantage appears only below roughly 140 vehicles, and the crossover is a central point of the paper's claims. Without statistical evidence, the differences shown in Fig. 7 could be within simulation noise.
[III.C, Eq. (3)] The score in Eq. (3) depends on a Delay value carried in the RREQ, but the manuscript does not specify how this delay is estimated during route discovery, which nodes update it, or how the destination obtains the values shown in Table I. The RREQ description says UAVs add only their IDs to the transited-zones list, so it is unclear how UAV-involved paths accumulate the delay metric. Without a defined delay estimator, the scoring rule is not reproducible.

minor comments (5)

[IV.B] The definition of average number of hops, 'the number of successfully delivered data packets divided by the total number of hops,' is the inverse of the intended quantity; the figure axes and the surrounding text indicate the metric is hops per delivered packet. Please correct the definition.
[IV.C] The phrase 'RBVT-R performs well than AGP' should read 'RBVT-R performs better than AGP.'
[III.C] The word 'undiserable' should be 'undesirable.'
[II] Reference [12] is cited as a 2016 publication for VanetMobiSim; please verify the correct year and version, as the mobility generator is widely known from earlier releases.
[IV.A] The simulation map is described only as 4000 m × 4000 m with 9 intersections; the road lengths, block sizes, and segment structure are not given, which makes it difficult to interpret the meaning of the zone sizes and vehicle densities.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular dependency: the protocol's scoring rule is defined independently of the reported simulation outcomes, and comparisons are against external baselines.

full rationale

The paper presents a reactive routing protocol with a fixed, explicitly stated path-scoring formula (Eq. 3) that combines measured RREQ information: NBvehicles, Delay, Sdeviation, and HOPs_UAV. No parameters were fitted to the NS-2 results, and no reported performance metric is defined as an output of that formula. The central claim, that cooperative UAVs extend connectivity in sparse urban VANETs, is evaluated by comparing the protocol against external baselines RBVT-R and AGP in simulation, rather than derived from the protocol's own assumptions. The only self-citation, reference [1] in the list [1]–[9], is used as contextual related-work support and is not load-bearing for any step of the proposed mechanism. The assumption in Section III.A that a UAV hovers over every four road segments is a design premise, not a quantity produced by the protocol, and any mismatch with the simulated UAV mobility is an experimental validity concern, not circularity. No self-definitional, fitted-input, self-citation-chain, or ansatz-smuggling reduction is present in the paper's derivation chain.

Assumptions & free parameters 3 free parameters · 4 assumptions · 1 invented entities

The protocol's performance depends on a small set of chosen parameters and assumptions about drone availability and wireless connectivity. The zone size and drone fleet size are hand-picked, and the coverage assumption is central to the claimed advantage.

free parameters (3)

Zone size = ≈300m
Zone size is set equal to the vehicle communication range; this choice determines the granularity of density recording and the path score, and no sensitivity analysis is provided.
Number of UAVs = 16
Chosen to satisfy the coverage assumption of at least one UAV per four road segments; no sensitivity analysis is reported.
Communication range = ≈300m
Standard assumption for vehicle-to-vehicle range, but it is the basis for zone sizing and affects connectivity, UAV placement, and hop counts.

assumptions (4)

domain assumption UAVs can communicate with vehicles through wireless interfaces up to a large transmission range and are not affected by obstacles.
Section III.A states this assumption; load-bearing because UAVs are expected to bridge gaps where ground connections fail.
domain assumption Each node has GPS, a map, and maintains an updated neighbor table.
Section III.A states this; required for zone-encoded path discovery and density counts.
domain assumption There is no energy constraint for vehicles or UAVs.
Section III.A states this; permits continuous UAV hovering and relaying, which the protocol depends on.
ad hoc to paper The network has a sufficient number of UAVs so that at each moment at least one UAV hovers an area of four road segments.
Section III.A; without this coverage guarantee, the protocol may not find a drone relay when the ground network is fragmented.

invented entities (1)

fixed zone
purpose: Partition road segments into spatial cells that carry vehicle density counts in route requests and path encoding.
A protocol abstraction, not a physical entity; its size is a free parameter and its utility is only demonstrated in simulation.

how reviews work

0 comments

Cite this review

Pith. "Pith review of On-Demand Routing for Urban VANETs using Cooperating UAVs." pith.science (2026). https://pith.science/paper/NTB7RK7E

@misc{pith2026190805905,
  author       = {Pith},
  title        = {Pith review of: On-Demand Routing for Urban VANETs using Cooperating UAVs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NTB7RK7E}},
  note         = {Machine review of arXiv:1908.05905}
}

read the original abstract

Vehicular ad hoc networks (VANETs) are characterized by frequent routing path failures due to the high mobility caused by the sudden changes of the direction of vehicles. The routing paths between two different vehicles should be established with this challenge in mind. Stability and connectedness are a mandatory condition to ensure a robust and reliable data delivery. The idea behind this work is to exploit a new reactive routing technique to provide regulated and well-connected routing paths. Unmanned Aerial Vehicles (UAVs) or what are referred to as drones can be both involved in the discovery process and be full members in these discovered paths in order to avoid possible disconnections on the ground when the network is sparsely connected. The different tests of this technique are performed based on NS-2 simulator and the outcomes are compared with those of related on-demand routing protocols dedicated for VANETs. Interesting results are distinguished showing a reduced end-to-end delay and a high delivery ratio, which proving that this heterogeneous communication between vehicles and UAVs is able to extend the network connectivity.

Figures

Figures reproduced from arXiv: 1908.05905 by the authors.

**Figure 1.** Network topology. send packets geographically between a source and destination nodes. Moreover, they are also used in the path maintenance process. 1) The route request (RREQ) packet format: Several fields compose the RREQ packet (c.f., [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗

**Figure 3.** Path discovery process. C. Path selection The selection process is carried out only when the discovery process has finished by discovering at least two routing paths. Indeed, several metrics are calculated for every discovered path based on the received information through the RREQs. As shown in [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗

**Figure 4.** Path selection process. 1) The route reply (RREP) packet format: Two information are added by the destination to the RREP packet (c.f., [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗

Figures from the paper (3 more)

**Figure 5.** Figure 5: The route reply (RREP) format. D. Path maintenance Once the source receives the RREP packet, it will copy the discovered paths field into the header of the data packet and starts to send the data packet through the selected path. When the transited path disconnects, th…

**Figure 6.** Figure 6: Path maintenance process. an alternative path is selected from the available paths stored the header of data packet. After this selection, the UAV is designated as the next forwarder to transmit the data packet through this new path. It is worth noting that, the UAVs c…

**Figure 7.** Figure 7: Simulation results destinations which is not the case of AGP. AGP selects paths with a high density of vehicles independently of the number of intersections. As shown in Fig. (7d), RBVT-R generates less overhead packets in high density because it does not generate freq…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

13 extracted references · 13 canonical work pages

[1]

Intelligent UA V-Assisted Routing Protocol for Urban V ANETs,

O. S. Oubbati, A. Lakas, F. Zhou, M. G ¨unes ¸, N. Lagraa, and M. B. Yagoubi, “Intelligent UA V-Assisted Routing Protocol for Urban V ANETs,”Computer communications , vol. 107, pp. 93–111, 2017

work page 2017
[2]

SC- GRP: SDN-enabled connectivity-aware geographical routing protocol of V ANETs for urban environment,

D. K. N. Venkatramana, S. B. Srikantaiah, and J. Moodabidri, “SC- GRP: SDN-enabled connectivity-aware geographical routing protocol of V ANETs for urban environment,”IET Networks, vol. 6, no. 5, pp. 102– 111, 2017

work page 2017
[3]

MoZo: a moving zone based routing protocol using pure V2V commu- nication in V ANETs,

D. Lin, J. Kang, A. Squicciarini, Y . Wu, S. Gurung, and O. Tonguz, “MoZo: a moving zone based routing protocol using pure V2V commu- nication in V ANETs,”IEEE Transactions on Mobile Computing , vol. 16, no. 5, pp. 1357–1370, 2017

work page 2017
[4]

V ANET routing on city roads using real-time vehicular trafﬁc information,

J. Nzouonta, N. Rajgure, G. Wang, and C. Borcea, “V ANET routing on city roads using real-time vehicular trafﬁc information,” IEEE Transac- tions on V ehicular Technology,, vol. 58, no. 7, pp. 3609–3626, 2009

work page 2009
[5]

Real-Time Intersection-Based Segment Aware Routing Algorithm for Urban Vehicular Networks,

Y . R. B. Al-Mayouf, N. F. Abdullah, O. A. Mahdi, S. Khan, M. Ismail, M. Guizani, and S. H. Ahmed, “Real-Time Intersection-Based Segment Aware Routing Algorithm for Urban Vehicular Networks,” IEEE Trans- actions on Intelligent Transportation Systems , 2018

work page 2018
[6]

A Trafﬁc Aware Segment-based Routing protocol for V ANETs in urban scenar- ios,

S. Khan, M. Alam, M. Fr ¨anzle, N. M ¨ullner, and Y . Chen, “A Trafﬁc Aware Segment-based Routing protocol for V ANETs in urban scenar- ios,” Computers & Electrical Engineering , vol. 68, pp. 447–462, 2018

work page 2018
[7]

AGP: an anchor-geography based routing protocol with mobility prediction for vanet in city scenarios,

S. Yan, X.-y. JIN, and S.-z. CHEN, “AGP: an anchor-geography based routing protocol with mobility prediction for vanet in city scenarios,” The Journal of China Universities of Posts and Telecommunications , vol. 18, pp. 112–117, 2011

work page 2011
[8]

Delay analysis and routing for two-dimensional V ANETs using carry-and-forward mechanism,

J. He, L. Cai, J. Pan, and P. Cheng, “Delay analysis and routing for two-dimensional V ANETs using carry-and-forward mechanism,” IEEE Transactions on Mobile Computing, vol. 16, no. 7, pp. 1830–1841, 2017

work page 2017

Show all 13 references

[9]

MURU: A multi-hop routing protocol for urban vehicular ad hoc networks,

Z. Mo, H. Zhu, K. Makki, and N. Pissinou, “MURU: A multi-hop routing protocol for urban vehicular ad hoc networks,” in Proceedings of the 3rd Annual IEEE International Conference on Mobile and Ubiquitous Systems: Networking & Services , 2006, pp. 1–8

2006
[10]

UA V assisted disruption tolerant routing,

M. Le, J.-S. Park, and M. Gerla, “UA V assisted disruption tolerant routing,” in Proceedings of IEEE Military Communications Conference (MILCOM), 2006, pp. 1–5

2006
[11]

Simulation and Comparative Analysis of Packet Delivery in Flying Ad Hoc Network (FANET) Using AODV,

A. V . Leonov, G. A. Litvinov, and E. V . Shcherba, “Simulation and Comparative Analysis of Packet Delivery in Flying Ad Hoc Network (FANET) Using AODV,” in Proceedings of the 19th International Con- ference of Young Specialists on Micro/Nanotechnologies and Electron Devices (...

2018
[12]

VanetMobiSim: generating realistic mobility patterns for V ANETs,

J. Haerri, M. Fiore, F. Fethi, and C. Bonnet, “VanetMobiSim: generating realistic mobility patterns for V ANETs,” 2016

2016
[13]

Mobi-sim: A simulation en- vironment for mobile wireless sensors network,

O. Mezghani and P. M. Abdellaoui, “Mobi-sim: A simulation en- vironment for mobile wireless sensors network,” in Proceedings of the International Conference on Control, Decision and Information Technologies (CoDIT). IEEE, 2016, pp. 192–197

2016

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

Intelligent UA V-Assisted Routing Protocol for Urban V ANETs,

O. S. Oubbati, A. Lakas, F. Zhou, M. G ¨unes ¸, N. Lagraa, and M. B. Yagoubi, “Intelligent UA V-Assisted Routing Protocol for Urban V ANETs,”Computer communications , vol. 107, pp. 93–111, 2017

work page 2017

[2] [2]

SC- GRP: SDN-enabled connectivity-aware geographical routing protocol of V ANETs for urban environment,

D. K. N. Venkatramana, S. B. Srikantaiah, and J. Moodabidri, “SC- GRP: SDN-enabled connectivity-aware geographical routing protocol of V ANETs for urban environment,”IET Networks, vol. 6, no. 5, pp. 102– 111, 2017

work page 2017

[3] [3]

MoZo: a moving zone based routing protocol using pure V2V commu- nication in V ANETs,

D. Lin, J. Kang, A. Squicciarini, Y . Wu, S. Gurung, and O. Tonguz, “MoZo: a moving zone based routing protocol using pure V2V commu- nication in V ANETs,”IEEE Transactions on Mobile Computing , vol. 16, no. 5, pp. 1357–1370, 2017

work page 2017

[4] [4]

V ANET routing on city roads using real-time vehicular trafﬁc information,

J. Nzouonta, N. Rajgure, G. Wang, and C. Borcea, “V ANET routing on city roads using real-time vehicular trafﬁc information,” IEEE Transac- tions on V ehicular Technology,, vol. 58, no. 7, pp. 3609–3626, 2009

work page 2009

[5] [5]

Real-Time Intersection-Based Segment Aware Routing Algorithm for Urban Vehicular Networks,

Y . R. B. Al-Mayouf, N. F. Abdullah, O. A. Mahdi, S. Khan, M. Ismail, M. Guizani, and S. H. Ahmed, “Real-Time Intersection-Based Segment Aware Routing Algorithm for Urban Vehicular Networks,” IEEE Trans- actions on Intelligent Transportation Systems , 2018

work page 2018

[6] [6]

A Trafﬁc Aware Segment-based Routing protocol for V ANETs in urban scenar- ios,

S. Khan, M. Alam, M. Fr ¨anzle, N. M ¨ullner, and Y . Chen, “A Trafﬁc Aware Segment-based Routing protocol for V ANETs in urban scenar- ios,” Computers & Electrical Engineering , vol. 68, pp. 447–462, 2018

work page 2018

[7] [7]

AGP: an anchor-geography based routing protocol with mobility prediction for vanet in city scenarios,

S. Yan, X.-y. JIN, and S.-z. CHEN, “AGP: an anchor-geography based routing protocol with mobility prediction for vanet in city scenarios,” The Journal of China Universities of Posts and Telecommunications , vol. 18, pp. 112–117, 2011

work page 2011

[8] [8]

Delay analysis and routing for two-dimensional V ANETs using carry-and-forward mechanism,

J. He, L. Cai, J. Pan, and P. Cheng, “Delay analysis and routing for two-dimensional V ANETs using carry-and-forward mechanism,” IEEE Transactions on Mobile Computing, vol. 16, no. 7, pp. 1830–1841, 2017

work page 2017