<?xml version="1.0" encoding="UTF-8" ?>
<?xml-stylesheet type="text/xsl" href="https://devzone.nordicsemi.com/cfs-file/__key/system/syndication/rss.xsl" media="screen"?><rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:slash="http://purl.org/rss/1.0/modules/slash/" xmlns:wfw="http://wellformedweb.org/CommentAPI/" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Maximum value for ATT_MTU size for 10ms connection interval</title><link>https://devzone.nordicsemi.com/f/nordic-q-a/24709/maximum-value-for-att_mtu-size-for-10ms-connection-interval</link><description>I&amp;#39;m transmitting data packets each of 48 bytes in the form of notifications using Bluetooth 5.0. When I observed the logs over sniffer, I see that the packets are fragmented which is shown as CRC errors and L2CAP fragmentation. 
 Goal: To transmit and</description><dc:language>en-US</dc:language><generator>Telligent Community 13</generator><lastBuildDate>Wed, 30 Aug 2017 12:31:47 GMT</lastBuildDate><atom:link rel="self" type="application/rss+xml" href="https://devzone.nordicsemi.com/f/nordic-q-a/24709/maximum-value-for-att_mtu-size-for-10ms-connection-interval" /><item><title>RE: Maximum value for ATT_MTU size for 10ms connection interval</title><link>https://devzone.nordicsemi.com/thread/97256?ContentTypeID=1</link><pubDate>Wed, 30 Aug 2017 12:31:47 GMT</pubDate><guid isPermaLink="false">137ad170-7792-4731-bb38-c0d22fbe4515:345d852b-2ad4-4df1-a016-db5f4030d235</guid><dc:creator>Nivetha</dc:creator><description>&lt;p&gt;Thank you so much, Peter. I have set 48 bytes as minimum data size with ATT_MTU of 52 bytes so that I can send with a very low/0 latency. However, my doubt is how far can I increase the ATT_MTU &amp;gt;52 bytes for 10ms connection interval such that I can stuff more data in with low latency and achieve high throughput. A Best optimal solution. I think I need to do more testing to come to a trade-off. I have also added DLE (ATT_MTU size + 4 bytes) at the central and peripheral.&lt;/p&gt;&lt;div style="clear:both;"&gt;&lt;/div&gt;</description></item><item><title>RE: Maximum value for ATT_MTU size for 10ms connection interval</title><link>https://devzone.nordicsemi.com/thread/97255?ContentTypeID=1</link><pubDate>Wed, 30 Aug 2017 12:17:48 GMT</pubDate><guid isPermaLink="false">137ad170-7792-4731-bb38-c0d22fbe4515:641eba91-1468-4bd7-b88d-311eb64d4aeb</guid><dc:creator>Petter Myhre</dc:creator><description>&lt;blockquote&gt;
&lt;p&gt;How to determine the max att_mtu size
for this connection interval such that
there is very less/no delay and high
throughput?&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;The latency (I guess this is what you mean by delay) is mostly determined by the connection interval. While the throughput depends on a lot of things. What kind of throughput do you want? 48 bytes each way with a connection interval of 10 ms? This isn&amp;#39;t really high throughput, and shouldn&amp;#39;t be a problem.&lt;/p&gt;
&lt;blockquote&gt;
&lt;p&gt;Does DLE extension effect this delay
and high throughput? if so, then what
should be the ATT_MTU size with and
without DLE?&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;DLE extends the maximum size of the link layer packets, while increasing the ATT_MTU increases the higher layer attribute packets. Increasing the ATT_MTU doesn&amp;#39;t really help that much if you don&amp;#39;t do DLE as well. If you don&amp;#39;t use DLE the ATT packets will be fragmented, sent over the air, and then reassembled on the other side of the link. DLE in general gives you higher throughput (as long as you don&amp;#39;t have a lot of packet loss), doesn&amp;#39;t affect latency.&lt;/p&gt;
&lt;blockquote&gt;
&lt;p&gt;Also, I added the battery service
which sends data every 1 sec. So how
to adjust the ATT_MTU size
accordingly?&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;I don&amp;#39;t understand. Why would you need to adjust the ATT_MTU size for this?&lt;/p&gt;
&lt;blockquote&gt;
&lt;p&gt;Does the DLE and connection event
extension be set in both central and
peripheral?&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;DLE is a symmetric procedure, so yes. It doesn&amp;#39;t help if your device can send a large link layer packet if the peer can&amp;#39;t receive it. Connection event extension should usually be set on both sides yes, but be aware that this is a Nordic specific thing, while DLE is BLE spec.&lt;/p&gt;&lt;div style="clear:both;"&gt;&lt;/div&gt;</description></item></channel></rss>