Multitier Fog Computing With Large-Scale IoT Data Analytics for Smart Cities
read more
Citations
All one needs to know about fog computing and related edge computing paradigms: A complete survey
Blockchain for Internet of Things: A Survey
Edge computing: A survey
All One Needs to Know about Fog Computing and Related Edge Computing Paradigms: A Complete Survey
Narrow Band Internet of Things
References
A view of cloud computing
Spark: cluster computing with working sets
Fog computing and its role in the internet of things
Internet of Things for Smart Cities
Resilient distributed datasets: a fault-tolerant abstraction for in-memory cluster computing
Related Papers (5)
Frequently Asked Questions (16)
Q2. What is the role of the resource module in fog?
2) Networking and Virtualization Module: Networking is a critical part of fog, especially for the scenarios where adhoc fogs nodes are mobile and the wireless link bandwidth is limited.
Q3. What is the main concern of the data analytics service users?
For the data analytics service and the QoS aware job management schemes, the major concern from the data analytics service users is on the service quality such as job blocking probability; while for the service operators of the fogs and the clouds they are more concerned on the utility (or revenue) generated from the analytics services.
Q4. What are the main features of the proposed QoS aware resource management schemes?
QoS aware job admission control, offloading and resource allocation schemes were designed and developed to provide QoS support for large scale data analytics services over multi-tier fogs.
Q5. What is the blocking probability of a cloud computing architecture?
When the multi-tier computing architecture is used (with A-Fog, D-Fog and cloud), the overall analytics service has the lowest blocking probability, which is less than 0.03 for most job arrival rates.
Q6. What are the main dimensions of diversities for the benchmarking of analytics systems?
For the benchmarking of analytics systems there could be three major dimensions of diversities that need to consider [10]: analytics computing platform, analytics algorithm and analytics job dataset.
Q7. What is the impact of the desktop computer on the analytics service capacity?
In addition the desktop computer shows a large impact on the analytics service capacity, which substantially reduces job completion time and extends the size of datasets that can be processed.
Q8. What are the advantages of fog management?
Special incentive and reward schemes can be applied by fog masters to encourage interconnected devices to join fogs and share their unused computing resources.
Q9. Why does the blocking probability increase with job arrival rates larger than 0.2?
Due to the limited computing resources at the A-Fog, the blocking probability increases fast with job arrival rates larger than 0.2, which indicates strong need of dedicated computing resources at D-Fog and/or clouds.
Q10. How much is the cheapest mobile broadband data tariff in the UK?
According to their market research, the currently cheapest mobile broadband data tariff in the UK is 2 pounds per GB mobile data traffic.
Q11. How can the authors compute the computing cost for a given job?
Then the computing cost for a given job can be computed according to the computing resource allocated for this job, the computation time with the allocated resources, and the computing resource price.
Q12. What is the highest service utility for job arrival rates?
The multi-tier fog computing system achieves the highest servie utility for most job arrival rates, which is almost double of the cloud only system.
Q13. How many computer settings are used to evaluate and compare the analytics computing performance under different experiment configurations?
To evaluate and compare the analytics computing performance under different experiment configurations, 12 computer settings representing the number of CPU cores and the memory size are used, which are shownin Table.
Q14. What is the description of the multi-tier fog computing model?
In the multi-tier fog computing model there are both ad-hoc fogs with opportunistic computing resources and dedicated fogs with dedicated computing resources.
Q15. What is the blocking probability of a cloud only computing architecture?
On the other hand, the A-Fog only computing architecture can provide quite good service quality until job arrival rate is larger than 0.2 jobs per second.
Q16. What is the key to ad-hoc fog benchmarking?
in order to do so, a key is to measure and model the workloads of different analytics services over the fogs with various computing and communication resources.