What is Workload Modeling for LTE
Understanding the current subscriber behavior and trends is a valuable guide for a product manufacturer and operator of the capital investment. In this post a few key observations will be shared via the Internet and existing wireless high speed data transfer.
The first observation is:
1. Increased bandwidth translates into more demanding users.
This is not a new observation, but continues to be born. Higher speeds translate into more and more files to download with shorter thinking/reading. This trend is Follow the curve of the “power law” as follows:
GB per month = (Mbps connection speed) x 0.7 1.2
For example, dial-up users with link speeds of 64 Kbps will consume ~ 0.2 GB/month. An experienced user works on fiber link, providing 50 Mbps would consume 15 + GB/month. If the user’s bandwidth is approximately 0.5 -1.0 Mbps, load from 0.7 to 1.2 you can expect GB/month.
The second point is:
2. Daily data traffic distribution is flatter than voice traffic.
For those with cellular backgrounds was the rule of thumb suggests that approximately 10% of the daily load voice was present in the hour, busy. But, every day traffic distribution of data consistently show busy hour for ~ 7% of the daily volume or below. Note that the lower limit for this ratio is ~ 4% (= 1/24 for uniformly Distributed load).
Third point is:
3. Since 2000, the Internet became more symmetric among top-down and bottom-up and now about 55/45 in/out.
This observation is in comparison to the cellular data distribution, which is 80/20. IN the increase is attributable to the substantial increase in peer-to-peer (P2P) file sharing which represents 55% of the total traffic ~ and ~ 90% upward movement on Internet sharing.
P2P networks requires (and, to some extent, apply) download the joint international members of the community. Operators may try to block or hinder the top users (which are usually file sharing). Assume that without the influence of the top 4% Users, traffic distribution between the DL and UL will return to ~ 75/25.
What is Workload Modeling for LTE
Workload modeling in LTE is about predicting how users like you and me will use the network. It helps plan how much traffic will be on the system, so performance stays smooth.
For example, if you’re streaming video and I’m just browsing websites, our network usage is very different. Workload modeling looks at that difference and helps the network allocate resources accordingly.
It includes things like number of users, time of day, types of apps, and data rates. If most users in an area stream during evening hours, the model prepares the system for that spike.
This way, the LTE system doesn’t get overloaded and can serve everyone better. Without it, you might see drops or buffering just because too many people are online at once.