# How to calculate the latency in vehicular network?

I need to calculate the delay in vehicular network V2V and V2I and I define it as the following:

Where L [Bit] the length of data and the transmission rate

where is the distance between vehicle `i` and the RSU `m` while `c` is the vehicle speed.

Is this the right way to calculate the latency?

• Homework? Off-topic. – Zac67 Jul 28 '19 at 5:52
• Otherwise, this question seriously lacks research. Voting to close. – Zac67 Jul 28 '19 at 8:17
• @Zac67 It is not homework, I'm reading this paper ieeexplore.ieee.org/document/6060155 and I need to check if this the right way to calculate the latency in V2X. – i_th Jul 28 '19 at 9:57

Is this the right way to calculate the latency?

No, and I don't know how one would support such an estimate.

Use of cellular data involves one set of calculations and direct-to-roadside involves another.

True latency calculations start at the sensor, that data is processed by an algorithm (unless it is to be transmitted directly) and then it is routed over an appropriate channel.

For example vehicle location and speed could be usefully sent both to roadside receivers and to nearby cellular towers. Not only would it offer redundancy but the manner in which the information would be used differs between the two endpoints.

Roadside receivers might make use of the speed and congestion data to set a variable speed limit which relied on local weather, while information sent to cell towers might be routed to servers which calculate travel time by various routes.

People already driving on the road don't need to know that if they had went a different way in the first place that they could have shaved a couple of minutes off of their travel time and people choosing a particular route probably don't care about local conditions if a particular route ends up being 10 minutes faster.

What data is collected, how it is processed before transmission, where it's going to and what it's destination is will all affect latency. If you want collision avoidance you want latency (and error free calculations) to be a top priority, while for congestion calculations you want an average of a 5 or 10 minute interval. You wouldn't want for people to be stopped at a light resulting in an overestimation, nor for a fender bender making it seem like the road was closed and an enormous speedup calculated for the cross-street.

In the paper "Data proxies, the cognitive layer, and application locality: enablers of cloud-connected vehicles and next-generation internet of things" (Jan 2016), on page 63, author Josh Siegal of MSU writes:

"Data freshness is complicated to evaluate, as communication is often either high power or high latency, and processing between hops can be significant – as much as 10s per 1km for V2V applications, assuming there are enough vehicles to ensure robust connectivity [94] [109] [116]. Power limitations and the potential for vehicular obstruction require adaptive communication to ensure sufficiently-low latency to enable safety applications [122]. Networks today, like 802.11P, are adequate for small systems (about 5 vehicles), but require additional intelligence to reduce transmission delay [94]. Problems of queue filling and similar are especially bad in dense environments, where time-critical message dissemination is not possible [32].

It is not only the network that contributes to these delays – decision on what data to collect, process, filter, and transmit can have significant impact, as can additions of layers designed to improve robustness and security [2] [109].".

On page 67:

"Radio technology is an important consideration for connected cars. Modern vehicles may have radios ranging from cellular connectivity for phone calls to mesh networking for V2V applications to WiFi or Bluetooth [134]. The radio technology chosen for use in a vehicle has many implications on application feasibility; each technology has different range, latency, bandwidth, and cost constraints, along with different market penetration. Additionally, radio systems have differing robustness to motion, line-of-sight obstruction, and antenna design [135]. Handoff mechanisms may be used to fuse communication across radio types seamlessly and with little computational overhead [136]. In spite of this, the cost of installing radio modules in a vehicle is not insignificant, so the design decision of what networks to access must be taken seriously.".

On page 78:

"Modern Connected Car applications have a plethora of inputs and outputs available for use. These applications must be designed with consideration for whether the vehicle needs data from external sources or if these data can be stored a priori. This decision of application locality involves exploring the use of thick/thin clients based on computation, storage, bandwidth needs, the optimal split of real-time and free (DSRC), long range real-time and costly (3G), long latency and free (WiFi), and the future trends in the cost of computation, storage, and communications. These decisions may often result in hybridized solutions, such as the use of local data buffers and last-mile distribution through DSRC or another low-cost solution [105]. The choice of communication method can have substantial impact on an application’s performance, as in the case of Miller’s 2007 sparsity simulation showing that with V2V, 10% of vehicles transmitting speed and location can produce similar results to V2I with only 10% of the bandwidth [155].".

References:

[2] M. Faezipour, M. Nourani, A. Saeed, and S. Addepalli, “Progress and challenges in intelligent vehicle area networks,” Communications of the ACM, vol. 55, no. 2, p. 90, February 2012.

[32] S. U. Eichler, “Performance evaluation of the IEEE 802.11 p WAVE communication standard,” in Vehicular Technology Conference, 2007. VTC-2007 Fall. 2007 IEEE 66th. IEEE, 2007, pp. 2199–2203.

[94] J. A. Misener, R. Sengupta, and H. Krishnan, “Cooperative collision warning: Enabling crash avoidance with wireless technology,” in 12th World Congress on ITS, vol. 3. Citeseer, 2005.

[105] A. Bazzi, B. M. Masini, and G. Pasolini, “V2V and v2r for cellular resources saving in vehicular applications,” in Wireless Communications and Networking Conference (WCNC), 2012 IEEE. IEEE, 2012, pp. 3199–3203.

[109] F. Kargl, P. Papadimitratos, L. Buttyan, M. Muter, E. Schoch, B. Wiedersheim, T. V. V. Thong, G. Calandriello, A. Held, and A. Kung, “Secure vehicular communication systems: implementation, performance, and research challenges,” Communications Magazine, IEEE, vol. 46, no. 11, pp. 110–118, 2008.

[116] J. Santa, A. F. Gómez-Skarmeta, and M. Sánchez-Artigas, “Architecture and evaluation of a unified V2V and V2I communication system based on cellular networks,” Computer Communications, vol. 31, no. 12, pp. 2850–2861, 2008.

[122] M. Sepulcre and J. Gozalvez, “Experimental evaluation of cooperative active safety applications based on V2V communications,” in Proceedings of the ninth ACM international workshop on Vehicular inter-networking, systems, and applications. ACM, 2012, pp. 13–20.

[134] (2012, September) Delphi & Michigan Department of Trans- portation. [Online]. Available: https://www.michigan.gov/documents/mdot/09-27-2012_Connected_Vehicle_Technology_-_Industry_Delphi_Study_401329_7.pdf

[135] Z. H. Mir and F. Filali, “LTE and IEEE 802.11 p for vehicular networking: a performance evaluation,” EURASIP Journal on Wireless Communications and Networking, vol. 2014, no. 1, pp. 1–15, 2014.

[136] J. A. Olivera, I. Cortázar, C. Pinart, A. L. Santos, and I. Lequerica, “VANBA: a simple handover mechanism for transparent, always-on V2V communications,” in Vehicular Technology Conference, 2009. VTC Spring 2009. IEEE 69th. IEEE, 2009, pp. 1–5.

[155] J. Miller, “Freesim–a free real-time V2V and V2I freeway traffic simulator,” IEEE Intelligent Transportation Systems Society Newsletter, 2007.

An article on the website Fierce Wireless: "3G/4G wireless network latency: How do Verizon, AT&T, Sprint and T-Mobile compare?" has a chart derived from information obtained from the OpenSignal website, their article: "State of Mobile USA: Quantifying the bar for 5G to beat" which offers much more recent data:

Different cities have different carriers and different equipment, with a difference in latency of 50%.

Different cities around the world rely on 2G-5G, that affects latency too.

• Interesting read, but the article the OP linked to is talking about "A V2X-based approach for reduction of delay propagation in vehicular Ad-Hoc networks". – Zac67 Jul 28 '19 at 15:03
• That's how they are implemented, and what the first linked (and first bold characters) refer to. – Rob Jul 28 '19 at 15:15
• Many thanks for your answer. As I read many papers and there are many ways. So, I don't know which one to use and this way I found it here ieeexplore.ieee.org/document/6060155 and this paper published in May 2019 used this way ibb.co/JmMQSRN – i_th Jul 28 '19 at 15:31
• You'll need to add acquisition and processing time to the averages experienced by people using different systems (shown in the charts at the end of my answer). Radio waves travel at almost the speed of light, the formula in your question includes the vehicle speed ? but the variable c is usually the speed of light. With so many vehicles and so much data the radio packet collision is bound to be the largest number. – Rob Jul 28 '19 at 15:54

Generally, the total delay consists of

• propagation delay (distance/signal velocity; speed of light for wireless)
• serialization delay (data length/transmission rate)
• processing delay (depending on hardware/software)
• Thanks for your reply. I asked this question because I found different ways to calculate it. See this ibb.co/JmMQSRN and I'm not sure which is the one is the right way. – i_th Jul 28 '19 at 13:00
• @i_th That is about LTE that requires an additional handshake - various protocols have various overheads and according inherent delays. I had the impression you were asking generally. – Zac67 Jul 28 '19 at 13:44

I need to share what I found about the delay

Transmission delay= Packet size/ transmission rate =L/R

Propagation Delay = Distance / Speed = D /S

processing delay can be ignored unless if the processing delay is quite large

Average Queuing delay = (N-1)L/(2*R)

where N = no. of packets
L=size of packet R=transmission rate

Paranjothi, A., Tanik, U., Wang, Y., & Khan, M. S. (2019). Hybrid‐Vehfog: A robust approach for reliable dissemination of critical messages in connected vehicles. Transactions on Emerging Telecommunications Technologies, e3595. http://zahid-stanikzai.com/types-of-delay/ https://www.geeksforgeeks.org/computer-network-packet-switching-delays/

In this paper:

Mostafa, A., Vegni, A. M., Singoria, R., Oliveira, T., Little, T. D., & Agrawal, D. P. (2011, August). A V2X-based approach for reduction of delay propagation in vehicular Ad-Hoc networks. In 2011 11th International Conference on ITS Telecommunications (pp. 756-761). IEEE.

The DELAY is

Where L [Bit] the length of data and the transmission rate and

where is the distance between vehicle `i` and the vehicle `j` while `c` is the vehicles traveling speed.

As well as, it can be defined as the following:

In this paper:

Yang, H., Xie, X., & Kadoch, M. (2019). Intelligent Resource Management Based on Reinforcement Learning for Ultra-Reliable and Low-Latency IoV Communication Networks. IEEE Transactions on Vehicular Technology, 68(5), 4157-4169.

The packet transmission delay is calculated as follows

where is the time needs to finish the three-way handshake process in the media access control (MAC) protocol and is the time needs to successfully complete one packet transmission