Depending on the topology/data-rate of your application, the radio becomes the significant power user. For large networks that monitor and control the protocol will effect the power consumption to a larger degree. There are companys that have been developing these protocols for over ten years. For example: www.millennialnet.com
The Cortus APS3 processor may be worth a look (www.cortus.com). It offers low power (22 µW/MHz in 130 nm), small size (similar footprint to an 8 bit core) and is powerful (1.15 DMIPS/MHz).
A 32 bit core like APS3 should also require less cost in software development than an 8 bit core.
recently, I am choosing one controler for GPS tracker, it should be low power,small size,and powerfull, so cortex M0 is better?
one similiar solution link:
The internet of things is being driven from many quarters. One primary one is the ubiquitousness of networking and low cost processing allowing things to talk to things such that humans get the distilled answer or final status. Why get 10000 sensor inputs from the Golden Gate Bridge when a user can just get the cloud computed stress levels at various parts of the span. This environment has room for many different protocols based on the M2M application.
@patk, when its M2M you can keep the type of traffic completely human unaccessible and application driven only. If its human readable information, then there are methods now that are suitable. Ethernet is the most widely used datacom network. Aside from physical security, its base security is not high. If you have access to your neighbours network you just plug in. Unless they run an authentication server, you are in. With 802.11 and 802.15.4 there is AES level security if users have enough concern to use it. This does not require the physical level protection. Though in a confined location (air base, for example, a limited accesss zone can also be used).
Even with authentication on ethernet, we have viruses and hackers getting in. That is a symptom of other levels of security that are breakable. Ultimately, the level of care must be appropriate for the desired protection. The technology already exists. Risk vs $$ and time. That's called insurance.