Re: MAC autentication of several packets in sensor network questions
- From: "Tom St Denis" <tomstdenis@xxxxxxxxx>
- Date: 30 Oct 2006 04:23:43 -0800
U Karlsson wrote:
The problem is: Authenticate packets from a sender (basestation(PC)) to
receiver (sensor-motes).
Constraints: The hardware on motes is 128 KB memory, and two 160 mAh
batteries, that shall work for an up to one year. So the number of
communication packets needs to be very low, computation can be little
higher.
The packet-size in the network is at least 20 B and most 128 B. So a
MAC shall include as little overhead as possibly.
Question: if the sender and receiver work as this, what needs to be
saved at the receiver and sender part? What is the problem? If any.
Sender: If the BS put in one MAC in the fifth packet (for 5 packets
(1-5)),
If this is a commercial venture [of a large nature] you may want to
look into an AES hardware core of some sort. It takes less energy to
process AES in an ASIC than it would in the typical embedded processor.
Receiver: after each packet save the 128 bit value from the encrypt and
XOR, and continues with it then receiving next packet.
You can aggregate your MACs provided you do not act upon the messages
until they have all been authenticated.
Tom
.
- References:
- MAC autentication of several packets in sensor network questions
- From: U Karlsson
- MAC autentication of several packets in sensor network questions
- Prev by Date: desperate ??
- Next by Date: Re: AES use of constant
- Previous by thread: MAC autentication of several packets in sensor network questions
- Next by thread: Re: MAC autentication of several packets in sensor network questions
- Index(es):
Relevant Pages
|
|
