QoS prediction is used to find out missing QoS values for services that are not by user and service selection is used to select the service with best QoS value.
is there any difference these terms or both are same?
please help....
QoS prediction is based in many parameters such as delay, rss, distance, path loss ,bw, channel availability and capacity also.
but prediction of cloud. services are: Infrastructure-as-a-Service (IaaS), Platform-as-a-Service (PaaS), and Software-as- a-Service (SaaS)
The approaches for cloud service selection can be categorized as: item-based approaches , customer-based approaches, their fusion approaches , model-based approaches , and ranking-oriented approaches. These approaches help to predict the missing values for QoS such as market dynamics, customer preferences etc.
According to my understanding, Quality-of-Service (QoS) prediction is an essential technique to select suitable cloud services. There is a wide range of literature available on predicting QoS value for cloud services. Following are a few examples for you to refer:
1. Article Multi-valued collaborative QoS prediction for cloud service ...
2. Article Time-aware cloud service recommendation using similarity-enh...
I hope this may help you. Thanks
Service selection is primarily employed to minimize operational costs while meeting a SLA. QoS is a KPI for an SLA, but not the only consideration.
Without the SLA, you have no motivation not to employ the cheapest possible compute assets. There is also the perceptual / human aspect. Every customer may have an SLA, or the idea of a SLA maintained on their behalf, however their ability to notice violations of the SLA, let alone exercise financial penalty clauses varies greatly. Then you have second order human elements: will the customer walk, will they trash you in social media, will they do one then the other.
So I would argue that QoS, as a composite or specific indicator, is a single input into the process that guides service selection.
I think the next layer to the optimization onion will be to automate managing the human / perceptual aspects of service delivery. As the current optimization methods are fairly robust when it comes to accounting for the gross financial impact of various modes of delivering service. But they are not forward looking.
In then end you need an algorithm that not only seeks to maximizes instantaneous profitability - but can take into account future profitability. In other words transition from tactical to strategic, from a 5 minute view of the future to a time span more aligned with the customer relationship.
-- Bob
- Badges
- Science method