Paper15:Adaptive QoS for Mobile Web Services through Cross-Layer Communication

Paper Title: Adaptive QoS for Mobile Web Services through Cross-Layer Communication

Three Critical Questions

Group 1:

Member Name: Bokka Satya Saradhi

1)How does the inclusion or augmentation of XML-schema in WS-QOS had lead
to efficient comparison of QOS requirements . What features of XML- Schema
properties made it to best suit for this technique?

2)Why do they complicate their system , by concentrating on throughput and efficiency of each layer of the system rather than concentrating on simply overall performance of the system?

3)In this system they have been relying on service provider , service broker very much . But how can they trust the reliability of them? , because they may tie up with the specific companies & may give variable results as best results rather than original ones that user had expected.

4)He had mentioned that BPEL4WS will take care of reliability , accessibility , security etc.. these are all comes under , non- functional requirements .Why does he ignored functional QOS parameters , which are far important ? .

Member Name:Vallurupalli Anusha

1)The author mentioned QOS with multiple parameters in the paper. Is controlling QOS with more than one parameter not a complex issue?

2)The author mentioned the interaction between different layers of the architecture as an important aspect for providing QOS. Are all the layers capable of operating in multiple modes of operation?

3) How the decompression technique causes almost negligible performance loss?

Group 2:

Member Name:Chiranjeevi Ashok Puvvula

Critical Questions:

1.What should the service provider do if the QOS filtering and QOS selection algorithms are ineffective and or not accurate? He is stuck to providing the desired QOS to the users?
2.Can the systems designed using cross layer architecture support dynamic QOS parameters? The paper completely discusses about static QOS parameters and no dynamic QOS parameter is specified.
3.Modular approach has really helped in standardization of interfaces and many methods. It was also proved that they really helped in fast pace development and research. SO, why modular approach can’t be implemented for achieving desired QOS as it has many advantages over cross layered architecture?

Group 2:

Member Name:Reena Rajan

Critical Questions:

1. The author says “We assume that the radio bearer supports both QoS and QoS classes as do the Universal Mobile Telecommunications Systems (UMTS) and the General Packet Radio service (GPRS) and that the wired network supports Diffserv technology”. What are the after effects if these conditions does not hold true? How will the proposed system work in such a situation?
2. The author says “…, processing the SOAP header leads to performance loss”. Are we willing to compromise on performance? What is more important performance or cross-layer QoS differentiation?
3. The author says, “Compression is one way to deal with Web services large message sizes”, “Depending on the current server load, the server compresses only the requests of those clients that required such a service”. Suppose if there are a number of clients that have requested for compression service and the server is overloaded. How will the system handle such a situation?

Group 3:

Member Name: Sharat Chandra Dammadapu

1. Since there is huge amount of data exchanged while using web services in mobile devices such as “HTTP and SOAP headers, XML data and brackets”. How would be the performance of sever vary? Why not JSON can be implemented instead of XML?
2. How security can be handled while integrating web services with mobile devices (wireless)? Since exchanging confidential data on a wireless device is not secured.
3. As mobile devices consists of constrained resources such as it is compatible with X GB RAM (memory), CPU. How the performance of web services varies?

Group 4:

Member Name:Srikanth Voruganti

Critical questions:
1)What happens if the client asks for some priorities and the WSB cannot match the
providers with those priorities? Will the WSB wait until the provider gives those priorities
or will generate errors to the clients about the situation?
2)The QoS metrics are changed from level to level during the communication. How well
are these metrics and techniques matched? Will there be any mechanism or set of
standards to ensure that the degree of the abstraction and security is maintained?
3)The parameters that affect the performance are interdependent. In the situation where a
web service meets most of the performance parameters and cannot deal with one of
the issue, what'll be the role of WSB and client to take the decision about the provider?

Group 5:

Member Name:SUNIL KUMAR GARREPALLY

1.It was mentioned that it uses compression techniques to reduce the size of SOAP messages to be transferred. What types of compression techniques are used? Does it lead to any data loss? Does it apply to typical operations such as bank transactions or military that won’t at all accept even a small amount of data?

2.Web service brokers takes requests from the client, first it searches for the match in the network proxy, if it does not match it searches in the UDDI registry. If it won’t get any results in even in the registry what methods does it follow to find the appropriate web service that meets QoS requirements of the client?

3.In the process of searching the web service the client need to produce interface key and Qos requirements to broker. What is interface key? What is the significance of it? How can it be generated?

4.As this process involves more steps (9 steps), it involves more messages to be transferred among client, broker & provider. Does it take any measures to control the traffic load on the network without affecting the data quality and data loss?

5.Very few parameters may have been defined in the system. If the mentioned QoS parameters were not defined, how the system recognizes these new parameters? Or does the system allow the user to define new QoS parameters based on the existing parameters?

Group 6:

Member Name : Debargh Acharya

1. The authors have stated that mobile devices are severely resource constrained as they are low in CPU processing, memory and battery life web services incur additional overheads if requests are made by SOAP. So compressing web services is required but what are the compressing techniques involved and how the compressing can be done optimally as it can also cause some overhead, these points were not discussed by the author.

2. During web service lookup by the client inside the UDDI registries of WSB the list of services are returned by UDDI that implement interface key. The indexing method followed or implemented is very important as it decides which particular web services are selected for the list. This has not been considered or elaborated by the author in this paper.

3. There is always a performance loss while processing the SOAP header during the evaluation of QoS information inside SOAP header. But how this performance loss can be evaluated or what are the steps to decrease it is still an open question.

Member Name:Bhargav Sandeep Ramayanam

1. In the paper the author mentioned that service providers will charge the mobile client by considering the aspect of volume rather than by the connection time. By this the same client can access the web service for long time at lower cost. By this the remaining clients cannot access that web service which is the severe drawback. Is it correct way of charging the client by considering the volume rather than the connection time?
2. In the paper the author mentioned a new role in the model which is the Web Service Broker. By introducing this role the Quality of service can be achieved and its accessing time can also be lowered. Is it secure to use the broker as an intermediate between the client application and the UDDI registry? Does the broker provide all the services that the author mentioned in the paper?
3. The author had mentioned that there are two types of Quality of Service specifications, which are custom specifications and the pre defined specifications. By introducing priorities to the custom specifications, does the processing time of the SOAP header get increased? Through this it takes more time to process the client request at the broker and it takes more time to process the broker response at the client.

Group 7:

Member Name: Rajanikanth Beesabathini

1. The WS-QoS SOAP message contains client’s requirements waiting for the service providing those requirements. All users may not have the knowledge regarding the technical terms. There may be tenderfoot who knows nothing regarding the throughput, jitter, and delay. So what kind of facilities that the Cross-Layer communication is offering to such layman?

2. Based on the Clients requirements, the WS-QoS searches for the service providers whose parameters suits with the customer requirements. In case of matching of more than one service providers, what kind of criteria that it will follow to select the service to the customer?

3. Is the information provided by the providers accurate enough? Based on which criteria the UDDI Registry accepts those parameters? Is there any method in the UDDI and WSD to know the accuracy of the claimed values of the Providers?

4. There are different cell phone manufacturing firms and offering several phone models with various features. All these phones may not have same capacity and features like battery backup, hardware CPU etc. How the service providers find those features for a specific cell phone in order to process those requests?

Member Name: Rameshwaram Karunapriya

Critical Questions:

1. The article explicates that the consumer's requirement of consistent service performance as necessitated is forcing the obligation for QoS stipulations in web services, does QoS help in providing the trustworthy service performance as required by the consumer?
2. It is explained that suppleness is being provided for the service producers in order to publish and renew all the service offers with several aspects of QoS. Are all the characteristics of QoS being met?
3. As the article discusses about cross layer communication, communication should be made possible among diverse layers. Are the layers interdependent in complying with the QoS supplied by the other layers?

Group 8:

Member Name:Ramya Devabhakthuni

Critical Questions:
1.The paper stated that the ‘WS-QoS’ message are passed over the network and each component along the communication path can evaluate this header. How can the security be achieved if the message contains sensitive data?
2.The paper stated that the Access Point is used to map the client side requirements to the server side by evaluating the QoS data in the SOAP header. Is this possible when the network grows and what happens when the Access Point experiences high traffic?
3.The Web Service Broker is added to the existing ‘interaction model’. This reduces the response time as the WSB analyses the service offers in advance. But what happens when the WSB does not have any information regarding the client request? What happens to the response time in this case?

Group 9:

Member Name: PRIYANKA KONERU

1. The author mentioned about using compression and decompression algorithms for efficient data transfer but he didn’t mention what type of compression and decompression to be used or the best algorithm for compression and decompression.
2. The author mentioned about compression and decompression of data while sending and receiving. Lossy compression may occur during data compression but the author didn’t mention any thing to protect the data from lossy compression.
3. The web service broker just checks whether the service provided by the service provider matches the user requirement or not. He may approach several service providers for the availability of services as required by the client. But the author didn’t mention whether the web service broker compares the services provided by the providers and provides the user with the best service.

Member Name : SRI HARSHA JASTI
1. The author mentioned about using compression and decompression algorithms for efficient data transfer. He didn’t mention whether both compression and decompression could be done using same algorithm or different algorithm.
2. It is mentioned the performance of CPU decreases due to data compression. An alternative method for improving the performance isn’t mentioned in the paper.
3. As the data traffic increases, the overhead task of the web service broker, server also increases. So how to reduce this traffic isn’t mentioned in the paper.

Group 10:

Member Name:Jonnalagadda, Harideep Reddy

1) What are those so called predefined parameters that enable efficient comparison of qos offers???How are their definitions extended????

2) Do the cross layer communications need any type of components to help in efficient communication??? Can any component with particular intelligence level help in this sort????

3) Is it not necessary to find effective compression and decompression techniques which consumes less CPU time and resources???? How can such compression techniques be developed????

Member name: Manasa chennamaneni
. Even compressing the web services slow down the server performance why the author preferred to compress and decompress the web services?
2. Why the modular approach is not given that importance for QoS requirements even it has some advantages over cross layer architecture?
3. Web service broker is an intermediate which helps in looking up the services in UDDI registries with in less time but to what extent it is helpful in satisfying all the QoS requirements?

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License