Time-Efficient way of sending image between 2 Raspberry Pi - serialization

I have to send a low resolution image (128x256) from one raspberry pi to another in the fastest way possible. What are some possible ways to approach this problem? Should I send bytes over I2C or set up some kind of LAN connection over which I could transmit (unidirectionally) images between the raspberrys using their IPs? A delay of less than 0.5-1 second is preferable.

Depending on your constraints and whether the two pis are networked, the Dat Protocol is fast, works over a LAN using multicast UDP, and will handle the serialization/deserialization for you.
See also try-dat

Related

Capture RAW data from Ethernet using Wireshark

I am new to Wireshark and capturing packets and all Stuff. Let me get it to the straight.
I have a hardware which outputs its data over Ethernet using a UDP Broadcast. I Can directly plug a Ethernet Cable to a In-line RJ-45 Coupler (attached to the hardware) and my PC Running Wireshark.
REQUIREMENTS : I need to Capture RAW Data which my hardware is broadcasting so that it can be given to other team so as to know the format in which it is providing for further post processing.
What I Did : Initially , I connected the Ethernet Cable from my home and Started capturing the packets which didn't make any sense to me.
Can you please point out if I am going in correct direction ? Sorry if its a very basic question, but raw data from the hardware is important for my further tasks....
As far as any software can understand a wire you will always get a packet. Between you (in front of a computer) and the cable in the in the RJ-45 jack sits a NIC (network interface controller, i.e. your network card).
Your Ethernet NIC will read the current on the cable (in manchester encoding for ethernet) and synchronize itself to any Ehternet traffic on that cable. What does "synchronizing" mane in there? In front of any Ehternet traffic come 64 alternate bits of 0s and 1s which are meant to synchronize the clocks on both communicating NICs. Without proper clock synchronization some data may be misinterpreted.
But why I am talking about clock synchronization? Because if you want the data as RAW as it is on the cable you will not get it. A NIC will never send any synchronization bit to the rest of the computer, therefore it is absolutely impossible to read exactly what is on the cable by using software.
On the other hand I find hard to believe you want the RAW data as RAW as that. After the synchronization bits come an Ethernet encapsulated packed. Yup, Ethernet uses packets. They're link layer packets (layer 2 in OSI).
And wireshark gives you exactly that (in most cases, see note at the end for two exceptions to this rule): every Ehternet packet that the NIC understands, manages to sync, and manages to read without collision is sent to the kernel and then read by wireshark. A cable has electrical interference and has no provision against collisions (it's just a piece of cooper!) therefore the NIC abstracts things like interferences and collisions.
I'll repeat it once more: After abstracting the synchronization bits, sender collisions (which turn the cable into one huge interference) and plain interferences; all that remains is a stream of packets, one after the other.
Extra Notes
NICs sometimes do ignore some Ethernet packets: packets that are not directed to their MAC. This can be changed by enabling promiscuous mode (available in most NICs). This is irrelevant for broadcast packets.
There are exception to the rule of wireshark getting all the traffic coming from the NIC:
If the traffic comes incredibly quick, wireshark may drop out of kernel schedule and not see some packets. It happens, nothign can be done about it.
If you listen on all interfaces (as opposed to selecting a single interface to listen at), wireshark will strip the Ethernet (or Wifi) headers. This is a wireshark hack needed to make output files uniform (and possible to be read by other applications).
TL;DR, wireshark output (pcap) is pretty much just the stream of packets that it got from the NIC, one after the other. That is as RAW as you can get with software.

Having difficulty sending small lwip packets immediately using the lwip API

I am creating a server on a ST Cortex M3 device. I am using the lwip API and FreeRTOS. All is working, but the response time is way off. I am currently using lwip 1.3.2 and FreeRTOS 7.3.
A single client connects to the server and must have some time-critical data sent frequently. These packets are on the order of 6 or so bytes. Other times, I am sending upwards of 20K.
The problem I am having is that these smaller packets seem to be taking forever to be sent. I assume this is because lwip is waiting for more data to be enqueued to make more efficient transmissions. I cannot wait around for 2 or 3 seconds for the data to be sent; the client is expecting the data nominally in a few micro-seconds or milli-seconds.
I have tried using lwip_send and lwip_write. (I understand that one is the same as the other with a flag passed at the end. Just had to try...) I have tried setting TCP_NODELAY on the socket to no avail. I tried to set SO_SNDLOWAT to '1', but this always returned -1, so I do not think it is supported.
I do not want to redo all of my code using TCP RAW. Is there a way to invoke the tcp_output() function outside of TCP RAW mode? Is there any way to speed things up or is this just how slow lwip TCP with small packets is?
Any and all suggestions are welcome. Thanks.
--EDIT--
I would also like to add that once I am ready to transmit, I make sure that my TX task in FreeRTOS is at the highest priority. There are no other tasks running up to the point at which I call lwip_send/write.
I'm fairly experienced with bare metal lwIP on xilinx and lwip does not wait to send things out. It will pump packets out as fast as your interrupts are acknowledged based on the ethernet hardware. I've been using UDP only. What is coming to mind though, is your problem might be on the receive end. If you are doing TCP, maybe those small packets are coming out late because you are having receive issues. What you need to do is find in the code the lowest level point at which ethernet is transmit, put a general purpose output toggle on that. Then also put a general purpose output toggle on when a ethernet packet is received. Look at the signals on a scope. If it confirms your hypothesis, then move the output toggles around to narrow down the issue. Wash, rinse and repeat until you are down to where the issue its. It's crude and time consuming, but oftentimes this brute force approach solves many "impossible" embedded software problems, due to pure determination. Good luck!

Using WinPCap for UDP receiving

I would like to use WinPCap library for "reliable" UDP receiving in my C++ application. All examples that I found, using this library for capturing and then proceding. Is there any way (example) how to configure PCap for streaming mode and receive UDP only and on uder defined port or how to solve this. In this time I have reliable UDP server able to receiving 0.5Gb/s. But on slower PC I have a packet lose I can see packets in ethereal but not in application.
thanks
vsm
I assume that you have already tried all of the more standard methods of increasing the number of datagrams that you are able to process? Things like increasing the recv buffer size, speeding up the processing that you do per datagram and using IOCP to allow you to bring more threads to bear on the problem or using RIO if you can target Windows 8?
If so then using WinPCap might work but it sounds like a bit of an extreme solution.
What you need to do is create a filter so that you only capture the datagrams that you are interested in... The docs include examples which use filters.
I have server from here: http://www.gamedev.net/topic/533159-article-using-udp-with-iocp/. This code working with IOCP. Its working fine on WIndows XP. There is no problem with receiving 0.5Gb/s. But now on Win7 is little unreliable. Sometimes there are packets positions error. (my device generating udp packets and in its payload there is PacketNumber - number increasing with each packet. When error occured i write all packet numbers into file. I can see for exmaple: 10,11,290,13,14... ). Is there any known differences in WinXP and Win7 for IOCP and multi threading? Or do you konw any free UDP server with IOCP processing?
In procedding loop I only adding packets into buffer and checking their numbers.

Throttling multicast datagrams

I have an application that is sending some UDP packets using multicast. I looked at the network traffic and there seems to be a lot of ancillary packets related to using multicast. I don't totally understand it, but does multicast by nature result in MORE network traffic. If so how can I throttle this down?
x
Other than the Multicast group join/remove messages, there are no ancillary messages created from you sending multicast data.
However, NIC's, routers, switches, printers, etc. all usually send some kind of multicast traffic, which is probably what you are seeing if you record the traffic.
In short, you need the networking equipment that forwards traffic between the client nodes to take care of this. Those vary depending on the network topology but would normally be:
Ethernet switches
IP routers.
Switch / router (implements functionality of a switch & router)
There are multicast control protocols such as IGMP but of course the source nodes and/or intermediate nodes (e.g. switches) must comply to these control protocols.
And YES multicast result in more network traffic : this is why plain Ethernet hubbing is practically extinct and additions to IEEE Ethernet such as VLANs are prevalent nowadays.
This is probably best addressed on some other sites (maybe this SO-style site PacketDrop).
LLC packets means you probably have sub-netting on your local segment, usually this doesn't mean extra packets though. You should change the network to a full class C if you want to remove LLC. On regular packets LLC or SNAP adds a 8-byte header.
http://ckp.made-it.com/ieee8022.html

For UDP broadcast gurus: Problems achieving high-bandwidth audio UDP broadcast over WiFi (802.11N and 802.11G)

I'm attempting to send multichannel audio over WiFi from one server to multiple client computers using UDP broadcast on a private network.
I'm using software called Pure Data, with a UDP broadcast tool called netsend~ and netreceive~. The code is here:
http://www.remu.fr/sound-delta/netsend~/
To cut a long story short, I'm able to achieve sending 9 channels to one client computer in a point-to-point network, but when I try to do broadcast to 2 clients (haven't yet tried more), I get no sound. I can compress the audio and send 4 channels compressed (about 10% size of uncompressed) over UDP broadcast to 2 clients successfully. Or I can send 1 channel over UDP broadcast to 2 clients, with some glitches.
The WiFi router is a Linksys WRT300N. All computers are running Windows XP. The IP addresses are 192.168.1.x, with subnet mask 255.255.255.0 and the subnet broadcast address: 192.168.1.255.
I'm curious - what happens to UDP broadcast packets in the router?
If I have a subnet mask of 255.255.255.0, then does the router make 254 packets for every packet sent ot the broadcast address?
My WiFi bandwidth is at least 100Mbps, but I can't seem to send audio of more than around 10Mbps over UDP broadcast to multiple clients.
What's stopping me from sending audio up to the bandwidth limit of the WiFi?
any suggestions for socket code modifications, network setups, router setups, subnet modifications... all very much appreciated!
thanks
Nick
Your problem is caused by the access point's rate control algorithm. With unicast the access point tracks what data rate every particular receiver can reliably receive at, and sends about that rate. With multicast the access point does not know which receivers are interested in the data, so simple access points send the data at the slowest possible rate (1Mb/s). Better implemented access points may send the data at the rate that the slowest connected client is using, and the best access points use IGMP snooping to see who's receiving each IP multicast stream, and they will choose the slowest rate out of the receivers for that stream.
The simplest solution is to not use multicast when you have a small number of WiFi receivers.
Are all parties connected via WiFi or is the sender using a
wired connection to the Access Point? Broadcast data will
be transmitted as unicast data from a station to an access
point and the access point will then retransmit the data
as broadcast/multicast traffic so it will use twice the
on-air bandwidth compared to when the sender sits on the
wired side of the AP.
When sending a unicast frame the AP will wait for an ACK
from the receiving station and it will retransmit the
frame until the ACK arrives (or it times out). Broadcast/multicast
frames are not ACKed and therefore not retransmitted.
If you have a busy/noisy radio environment this will
cause the likelyhood of dropped packets to increase,
potentially a lot, for multicast traffic compared to unicast
traffic. In an audio application this could certainly be audible.
Also, IIRC, broadcast/multicast traffic does not use the
RTS/CTS procedure for reserving the media which exarbates
the dropped packets problem.
It could actually be the case that multiple unicast streams
work better than a single multicast stream under less-than-ideal
radio conditions given that the aggregated bandwidth is
high enough.
If you can I would suggest that you use wireshark to sniff
the WiFi traffic and take a look at the destination address
in the 802.11 header. Then you can verify if the packets
are actually broadcast or not over the air.
Your design is failing due to a common misconception with WiFi speeds. With 802.11n the number 300mb/s is the link speed, not the actual bandwidth available for user data or even the IP layer. The effective bandwidth is closer to 40mb/s best case, have a look at the FAQ on SmallNetBuilder.com that discusses this in further detail.
http://www.smallnetbuilder.com/wireless/wireless-basics/31083-smallnetbuilders-wireless-faq-the-essentials
I'm curious - what happens to UDP broadcast packets in the router? If I have a subnet mask of 255.255.255.0, then does the router make 254 packets for every packet sent ot the broadcast address?
No the "router" doesn't make 254 individual packets. Furthermore, I suspect the protocol leverages "multicast" addresses rather than using a "broadcast" address.
Since broadcast/multicast traffic can easily be misused, there are many networking equipment that limit/block by default such traffic. Of course, some essential protocols (e.g. ARP, DHCP) rely on broadcast/multicast addresses to function and won't be blocked by default.
Hence, it might be a good thing to check for multicast/broadcast control knobs on your router.