Chapter 3 Design Patterns for IoT 73
One more complexity issue is associated with the absence of tools through which
developers can measure the latest applications. The standard ping and trace route which is
used for testing connections across the Internet is out-dated because the latest IoT networks
and applications do not have any link with ICMP. On the other hand, they require FTP, HTTP,
and similar protocols and hence you have to assess their performance accordingly.
By going over the Quality of Services (QoS) and prioritizing trac, you can go in more
depth of the complexity for the cloud latency. Before the advent of cloud, QoS and SLAs (Ser-
vice Level Agreements) were generated to ensure the prioritization of trac and make sure that
the applications which are sensitive for latency can have the appropriate network resources at
their disposal. The use of virtualized and cloud services is also obsolete; you have to distinguish
items like a network card, a piece of storage infrastructure, an outage in a server, or a security
risk. Cloud applications suer in terms of their fault tolerance for network latency.
Applications which manage back-oce reporting may accept diminished uptime; however,
there are many corporate processes where you cannot aord downtime or else the business suf-
fers major damage. Hence, it is necessary that SLAs are focused on certain applications on the
basis of availability and performance.
One of the most eective solutions is the connection to a cloud platform. For example,
Windows Azure is robust and powerful enough, especially for those companies which intend
to design hybrid IoT solutions where a portion of data is saved on the on-premises resources
while the cloud migration can factor in other components of the solution.
Lastly, you can collocate your IoT components in third party datacenters. This can assure
businesses that the cloud applications can manage all types of challenges related to latency while
they can also take advantage in the form of troubleshooting, support, cost, and monitoring.
Challenge #4: Quality of Service
It is normal for website users to get aected from changes in latency and meet occasional lapses in
web services. However, in the IoT spectrum, if a sensor is rendered unavailable for a short period
of time, then the repercussions can translate dangerously to the physical world. Considerable
engineering attempts have been poured to enhance the cloud latency and availability. The issue
is that these attempts have been hindered by DDoS attacks, standard variations in packets,
operation errors, bugs in the software, etc. Similarly, the connectivity against residential homes
is flawed. Almost 10% residential area networks in the developed countries struggle from
interruptions in connectivity. These hindrances occur once in 10days. Moreover, further issues
are increasingly identified in the developed world.
Solution
For mobile operators, the rising number of IoT devices forms a great prospect, an opportunity
through which they can benefit hugely. However, to achieve success with IoT applications, there
must be emphasis on trusted connectivity and reliability.
For example, Gemalto’s IoT QoS is a quite detailed solution through which mobile
operators can determine their cellular connectivity for mobile objects and smart static during
real time. The solution can address a wide range of usecases like smart metering, connected
PoS applications, smart home, and automotive IoT applications. What makes it unique is that
it oers quick analysis and status of the network. Moreover, it can rapidly identify issues by
making sure that a vast amount of data is always accessible.
Internet_of_Things_CH03_pp055-080.indd 73 9/3/2019 10:13:08 AM
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.